User equipment machine learning service continuity

Abstract

Aspects of the present disclosure generally relate to wireless communications. In some aspects, a user equipment (UE) can send first machine learning data associated with a machine learning service to a first network node for use by a first inference host associated with the first network node. The UE can receive, from the first network node, a handover command communication indicating that the UE is to perform a handover from the first network node to a second network node, where the handover command communication indicates machine learning inference information associated with a second inference host associated with the second network node. The UE can send second machine learning data for the machine learning service to the second network node for use by the second inference host associated with the second network node based at least in part on receiving the handover command communication. Numerous other aspects are described.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This patent application claims priority to U.S. non-provisional patent application No. 17 / 806,164, filed on June 9, 2022, entitled “USER EQUIPMENT MACHINELEARNING SERVICE CONTINUITY,” which is hereby expressly incorporated herein by reference. Technical Field

[0003] Aspects of the present disclosure relate generally to wireless communications, and to techniques and apparatus for user equipment (UE) machine learning service continuity. Background Art

[0004] Wireless communication systems are widely deployed to provide a variety of telecommunication services, such as telephony, video, data, messaging, and broadcasting. Typical wireless communication systems may employ multiple access technologies that can support communication with multiple users by sharing available system resources (e.g., bandwidth, transmit power, etc.). Examples of such multiple access technologies include code division multiple access (CDMA) systems, time division multiple access (TDMA) systems, frequency division multiple access (FDMA) systems, orthogonal frequency division multiple access (OFDMA) systems, single carrier frequency division multiple access (SC-FDMA) systems, time division synchronous code division multiple access (TD-SCDMA) systems, and long term evolution (LTE). LTE / LTE-Advanced is a set of enhancements to the Universal Mobile Telecommunications System (UMTS) mobile standard promulgated by the Third Generation Partnership Project (3GPP).

[0005] A wireless network may include one or more network nodes that support communications for wireless communication devices, such as a user equipment (UE) or multiple UEs. A UE may communicate with a network node via downlink communications and uplink communications. A "downlink" (or "DL") refers to a communication link from a network node to a UE, and an "uplink" (or "UL") refers to a communication link from a UE to a network node. Some wireless networks may support device-to-device communications, such as via a local link (e.g., a side link (SL), a wireless local area network (WLAN) link, and / or a wireless personal area network (WPAN) link, among other examples).

[0006] The above-mentioned multiple access technologies have been adopted in various telecommunication standards to provide a common protocol that enables different UEs to communicate at a city, country, region and / or global level. New Radio (NR) (which may be referred to as 5G) is a set of enhancements to the LTE mobile standard promulgated by 3GPP. NR is designed to improve spectrum efficiency, reduce costs, improve services, utilize new spectrum, and better integrate with other open standards by using orthogonal frequency division multiplexing (OFDM) (CP-OFDM) with cyclic prefix (CP) on the downlink, using CP-OFDM and / or single carrier frequency division multiplexing (SC-FDM) (also known as discrete Fourier transform spread OFDM (DFT-s-OFDM)) on the uplink, and supporting beamforming, multiple input multiple output (MIMO) antenna technology and carrier aggregation, so as to better support mobile broadband Internet access. As the demand for mobile broadband access continues to increase, further improvements in LTE, NR and other radio access technologies remain useful. Summary of the invention

[0007] Some aspects described herein relate to a user equipment (UE) for wireless communication. The UE may include a memory and one or more processors coupled to the memory. The one or more processors may be configured to send first machine learning data associated with a machine learning service to a first network node for use by a first inference host associated with the first network node. The one or more processors may be configured to receive a handover command communication from the first network node, the handover command communication indicating that the UE will perform a handover from the first network node to a second network node, wherein the handover command communication indicates machine learning inference information associated with a second inference host associated with the second network node. The one or more processors may be configured to send second machine learning data associated with the machine learning service to the second network node for use by the second inference host associated with the second network node, based at least in part on receiving the handover command communication.

[0008] Some aspects described herein relate to a first network node for wireless communication. The first network node may include a memory and one or more processors coupled to the memory. The one or more processors may be configured to receive a first handover communication from a first inference host associated with the first network node, the first handover communication indicating first machine learning inference information associated with a machine learning service, the machine learning service being associated with machine learning data associated with a first UE. The one or more processors may be configured to send a handover request communication associated with a handover process for the first UE to a second network node or a core network node, wherein the first network node selects the second network node for the handover process based at least in part on the first machine learning inference information and the machine learning service associated with the second network node. The one or more processors may be configured to receive a handover response communication associated with the handover process from the second network node or the core network node, the handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node.

[0009] Some aspects described herein relate to a first network node for wireless communication. The first network node may include a memory and one or more processors coupled to the memory. The one or more processors may be configured to receive a handover request communication associated with a handover process for a UE from a second network node or a core network node, wherein the handover request communication indicates first machine learning inference information associated with a machine learning service and machine learning data associated with the UE. The one or more processors may be configured to communicate with a first inference host associated with the first network node to determine that the first inference host is capable of supporting the machine learning service. The one or more processors may be configured to send a handover response communication indicating second machine learning inference information associated with the first inference host to the second network node or the core network node based at least in part on the first inference host being able to support the machine learning service.

[0010] Some aspects described herein relate to a core network node for wireless communications. The core network node may include a memory and one or more processors coupled to the memory. The one or more processors may be configured to receive a first handover request communication associated with a handover process for a UE from a first network node, wherein the first handover request communication indicates at least one of first machine learning inference information associated with a machine learning service, machine learning data associated with the UE, or a first inference host associated with the first network node. The one or more processors may be configured to send a second handover request communication associated with the handover process for the UE to a second network node, wherein the second network node is selected for the handover process based at least in part on the first machine learning inference information and the machine learning capability associated with the second network node. The one or more processors may be configured to receive a first handover response communication associated with the handover process from the second network node, the first handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node.

[0011] Some aspects described herein relate to a method of wireless communication performed by a UE. The method may include sending first machine learning data associated with a machine learning service to a first network node for use by a first inference host associated with the first network node. The method may include receiving a handover command communication from the first network node, the handover command communication indicating that the UE is to perform a handover from the first network node to a second network node, wherein the handover command communication indicates machine learning inference information associated with a second inference host associated with the second network node. The method may include sending second machine learning data associated with the machine learning service to the second network node for use by the second inference host associated with the second network node based at least in part on receiving the handover command communication.

[0012] Some aspects described herein relate to a method of wireless communication performed by a first network node. The method may include receiving a first handover communication from a first inference host associated with the first network node, the first handover communication indicating first machine learning inference information associated with a machine learning service, the machine learning service associated with machine learning data associated with a first UE. The method may include sending a handover request communication associated with a handover process for the first UE to a second network node or a core network node, wherein the first network node selects the second network node for the handover process based at least in part on the first machine learning inference information and the machine learning service associated with the second network node. The method may include receiving a handover response communication associated with the handover process from the second network node or the core network node, the handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node.

[0013] Some aspects described herein relate to a method of wireless communication performed by a first network node. The method may include receiving a handover request communication associated with a handover process for a UE from a second network node or a core network node, wherein the handover request communication indicates first machine learning inference information associated with a machine learning service and machine learning data associated with the UE. The method may include communicating with a first inference host associated with the first network node to determine that the first inference host is capable of supporting the machine learning service. The method may include sending a handover response communication indicating second machine learning inference information associated with the first inference host to the second network node or the core network node based at least in part on the first inference host being able to support the machine learning service.

[0014] Some aspects described herein relate to a method of wireless communication performed by a core network node. The method may include receiving a first handover request communication associated with a handover process for a UE from a first network node, wherein the first handover request communication indicates at least one of first machine learning inference information associated with a machine learning service, machine learning data associated with the UE, or a first inference host associated with the first network node. The method may include sending a second handover request communication associated with the handover process for the UE to a second network node, wherein the second network node is selected for the handover process based at least in part on the first machine learning inference information and a machine learning capability associated with the second network node. The method may include receiving a first handover response communication associated with the handover process from the second network node, the first handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node.

[0015] Some aspects described herein relate to a non-transitory computer-readable medium storing an instruction set for wireless communication by a UE. The instruction set, when executed by one or more processors of the UE, may cause the UE to send first machine learning data associated with a machine learning service to a first network node for use by a first inference host associated with the first network node. The instruction set, when executed by one or more processors of the UE, may cause the UE to receive a handover command communication from the first network node, the handover command communication indicating that the UE will perform a handover from the first network node to a second network node, wherein the handover command communication indicates machine learning inference information associated with a second inference host associated with the second network node. The instruction set, when executed by one or more processors of the UE, may cause the UE to send second machine learning data associated with the machine learning service to the second network node for use by the second inference host associated with the second network node based at least in part on receiving the handover command communication.

[0016] Some aspects described herein relate to a non-transitory computer-readable medium storing an instruction set for wireless communication by a first network node. The instruction set, when executed by one or more processors of the first network node, can cause the first network node to receive a first handover communication from a first inference host associated with the first network node, the first handover communication indicating first machine learning inference information associated with a machine learning service, the machine learning service being associated with machine learning data associated with a first UE. The instruction set, when executed by one or more processors of the first network node, can cause the first network node to send a handover request communication associated with a handover process for the first UE to a second network node or a core network node, wherein the first network node selects the second network node for the handover process based at least in part on the first machine learning inference information and the machine learning service associated with the second network node. The instruction set, when executed by one or more processors of the first network node, can cause the first network node to receive a handover response communication associated with the handover process from the second network node or the core network node, the handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node.

[0017] Some aspects described herein relate to a non-transitory computer-readable medium storing an instruction set for wireless communication by a first network node. The instruction set, when executed by one or more processors of the first network node, can cause the first network node to receive a handover request communication associated with a handover process for a UE from a second network node or a core network node, wherein the handover request communication indicates first machine learning inference information associated with a machine learning service and machine learning data associated with the UE. The instruction set, when executed by one or more processors of the first network node, can cause the first network node to communicate with a first inference host associated with the first network node to determine that the first inference host can support the machine learning service. The instruction set, when executed by one or more processors of the first network node, can cause the first network node to send a handover response communication indicating second machine learning inference information associated with the first inference host to the second network node or the core network node based at least in part on the first inference host being able to support the machine learning service.

[0018] Some aspects described herein relate to a non-transitory computer-readable medium storing an instruction set for wireless communication by a core network node. The instruction set, when executed by one or more processors of the core network node, can cause the core network node to receive a first handover request communication associated with a handover process for a UE from a first network node, wherein the first handover request communication indicates at least one of first machine learning inference information associated with a machine learning service, machine learning data associated with the UE, or a first inference host associated with the first network node. The instruction set, when executed by one or more processors of the core network node, can cause the core network node to send a second handover request communication associated with the handover process for the UE to a second network node, wherein the second network node is selected for the handover process based at least in part on the first machine learning inference information and the machine learning capability associated with the second network node. The instruction set, when executed by one or more processors of the core network node, can cause the core network node to receive a first handover response communication associated with the handover process from the second network node, wherein the first handover response communication indicates second machine learning inference information associated with a second inference host associated with the second network node.

[0019] Some aspects described herein relate to an apparatus for wireless communication. The apparatus may include means for sending first machine learning data associated with a machine learning service to a first network node for use by a first inference host associated with the first network node. The apparatus may include means for receiving a handover command communication from the first network node, the handover command communication indicating that the UE is to perform a handover from the first network node to a second network node, wherein the handover command communication indicates machine learning inference information associated with a second inference host associated with the second network node. The apparatus may include means for sending second machine learning data associated with the machine learning service to the second network node for use by the second inference host associated with the second network node based at least in part on receiving the handover command communication.

[0020] Some aspects described herein relate to a first device for wireless communication. The first device may include a component for receiving a first handover communication from a first inference host associated with the first network node, the first handover communication indicating first machine learning inference information associated with a machine learning service, the machine learning service associated with machine learning data associated with a first UE. The first device may include a component for sending a handover request communication associated with a handover process for the first UE to a second network node or a core network node, wherein the first network node selects the second network node for the handover process based at least in part on the first machine learning inference information and the machine learning service associated with the second network node. The first device may include a component for receiving a handover response communication associated with the handover process from the second device or the core network node, the handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node.

[0021] Some aspects described herein relate to a first device for wireless communication. The first device may include a component for receiving a handover request communication associated with a handover process for a UE from a second device or a core network node, wherein the handover request communication indicates first machine learning inference information associated with a machine learning service and machine learning data associated with the UE. The first device may include a component for communicating with a first inference host associated with the first network node to determine that the first inference host is capable of supporting the machine learning service. The first device may include a component for sending a handover response communication indicating second machine learning inference information associated with the first inference host to the second device or the core network node based at least in part on the first inference host being able to support the machine learning service.

[0022] Some aspects described herein relate to an apparatus for wireless communication. The apparatus may include a component for receiving a first handover request communication associated with a handover procedure for a UE from a first network node, wherein the first handover request communication indicates at least one of first machine learning inference information associated with a machine learning service, machine learning data associated with the UE, or a first inference host associated with the first network node. The apparatus may include a component for sending a second handover request communication associated with the handover procedure for the UE to a second network node, wherein the second network node is selected for the handover procedure based at least in part on the first machine learning inference information and a machine learning capability associated with the second network node. The apparatus may include a component for receiving a first handover response communication associated with the handover procedure from the second network node, the first handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node.

[0023] The various aspects generally include methods, apparatus, systems, computer program products, non-transitory computer-readable media, user equipment, base stations, network entities, network nodes, wireless communication devices and / or processing systems as fully described herein with reference to the drawings and description and as illustrated in the drawings and description.

[0024] The features and technical advantages of the examples according to the present disclosure have been outlined quite extensively above in order to better understand the subsequent specific embodiments. Additional features and advantages will be described below. The disclosed concepts and specific examples can be easily used as a basis for modifying or designing other structures for the same purpose of achieving the present disclosure. Such equivalent constructions do not depart from the scope of protection of the appended claims. The characteristics of the concepts disclosed herein, both in terms of their organization and method of operation, and the associated advantages will be better understood by considering the following description in conjunction with the accompanying drawings. Each of the drawings is provided for the purpose of illustration and description, and not as a definition of the limitations of the claims.

[0025] Although various aspects are described in the present disclosure by illustrating some examples, it will be understood by those skilled in the art that such aspects can be implemented in many different arrangements and scenarios. The technology described herein can be implemented using different platform types, devices, systems, shapes, sizes and / or packaging arrangements. For example, some aspects can be implemented via integrated chip implementations or other devices based on non-module components (e.g., end-user devices, vehicles, communication equipment, computing equipment, industrial equipment, retail / shopping equipment, medical equipment and / or artificial intelligence devices). Various aspects can be implemented in chip-level components, modular components, non-modular components, non-chip-level components, device-level components and / or system-level components. The equipment incorporating the various aspects and features described may include additional components and features for implementing and practicing the various aspects claimed and described. For example, the transmission and reception of wireless signals may include one or more components (e.g., hardware components, including antennas, radio frequency (RF) chains, power amplifiers, modulators, buffers, processors, interleavers, adders and / or summers) for analog and digital purposes. The various aspects described herein are intended to be practiced in various devices, components, systems, distributed arrangements and / or end-user devices of various sizes, shapes and configurations. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to fully understand the above-mentioned features of the present disclosure, a more specific description of the invention briefly summarized above can be obtained by referring to various aspects (some of which are illustrated in the accompanying drawings). However, it should be noted that the accompanying drawings only illustrate certain typical aspects of the present disclosure and are therefore not to be considered as limiting the scope thereof, as the specification may admit of other equally effective aspects. The same reference numerals in different drawings may identify the same or similar elements.

[0027] Figure 1 is a diagram illustrating an example of a wireless network according to the present disclosure.

[0028] Figure 2 is a diagram illustrating an example of communication between a network node and a user equipment (UE) in a wireless network according to the present disclosure.

[0029] Figure 3 is a diagram illustrating an example decomposed base station architecture according to the present disclosure.

[0030] Figure 4 is a diagram illustrating an example architecture of a functional framework for radio access network (RAN) intelligence implemented through data collection according to the present disclosure.

[0031] Figure 5 is a diagram illustrating an example of utilizing data from sensors of a vehicle to improve wireless communications according to the present disclosure.

[0032] Figure 6 is a diagram illustrating an example of a system architecture for utilizing sensed information from a sensor device to improve wireless communications in accordance with various aspects of the present disclosure.

[0033] Figure 7 is a diagram illustrating an example of a handover procedure according to the present disclosure.

[0034] Figure 8 is a diagram illustrating an example associated with UE machine learning service continuity according to the present disclosure.

[0035] Fig. 9A and Fig. 9B is a diagram illustrating an example associated with UE machine learning service continuity according to the present disclosure.

[0036] Fig.10 is a diagram illustrating an example associated with UE machine learning service continuity according to the present disclosure.

[0037] Fig.11 is a diagram illustrating an example process performed, for example, by a UE according to the present disclosure.

[0038] Fig.12 is a diagram illustrating an example process performed, for example, by a first network node according to the present disclosure.

[0039] Fig.13 is a diagram illustrating an example process performed, for example, by a first network node according to the present disclosure.

[0040] Fig.14 is a diagram illustrating an example process performed, for example, by a core network node according to the present disclosure.

[0041] Figures 15 to 17 is a diagram of an example apparatus for wireless communications according to the present disclosure. DETAILED DESCRIPTION

[0042] The various aspects of the present disclosure are described more fully below with reference to the accompanying drawings. However, the present disclosure can be embodied in many different forms, and should not be interpreted as being limited to any specific structure or function presented throughout the present disclosure. Instead, these aspects are provided so that the present disclosure will be thorough and complete, and the scope of protection of the present disclosure will be fully conveyed to those skilled in the art. It should be understood by those skilled in the art that the scope of the present disclosure is intended to cover any aspect of the disclosure disclosed herein, whether it is implemented independently or in combination with any other aspect of the present disclosure. For example, any number of aspects set forth herein may be used to implement a device or practice method. In addition, the scope of the present disclosure is intended to cover such a device or method implemented using other structures, functions, or structures and functions in addition to or different from the various aspects of the disclosure set forth herein. It should be understood that any aspect of the disclosure disclosed herein may be embodied by one or more elements of the claims.

[0043] Several aspects of telecommunication systems will now be presented with reference to various devices and techniques. These devices and techniques will be described in the following detailed description and illustrated in the accompanying drawings by various blocks, modules, components, circuits, steps, processes, algorithms, etc. (collectively referred to as "elements"). These elements may be implemented using hardware, software, or a combination thereof. Whether such elements are implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system.

[0044] Although various aspects may be described herein using terms generally associated with 5G or new radio (NR) radio access technology (RAT), various aspects of the present disclosure may be applicable to other RATs, such as 3G RAT, 4G RAT, and / or post-5G RATs (e.g., 6G).

[0045] Figure 11 is a diagram illustrating an example of a wireless network 100 according to the present disclosure. The wireless network 100 may be a 5G (e.g., NR) network and / or a 4G (e.g., Long Term Evolution (LTE)) network or may include elements of a 5G (e.g., NR) network and / or elements of a 4G (e.g., Long Term Evolution (LTE)) network, etc. The wireless network 100 may include one or more network nodes 110 (shown as network node 110a, network node 110b, network node 110c, and network node 110d), one user equipment (UE) 120 or multiple UEs 120 (shown as UE 120a, UE 120b, UE 120c, UE 120d, and UE 120e), and / or other entities. The network node 110 is a network node that communicates with the UE 120. As shown in the figure, the network node 110 may include one or more network nodes. For example, the network node 110 may be a converged network node, which means that the converged network node is configured to utilize a radio protocol stack that is physically or logically integrated within a single radio access network (RAN) node (e.g., within a single device or unit). As another example, the network node 110 may be a decomposed network node (sometimes referred to as a decomposed base station), which means that the network node 110 is configured to utilize a protocol stack that is physically or logically distributed between two or more nodes (such as one or more central units (CUs), one or more distributed units (DUs), or one or more radio units (RUs)).

[0046] In some examples, the network node 110 is or includes a network node that communicates with the UE 120 via a radio access link, such as an RU. In some examples, the network node 110 is or includes a network node that communicates with other network nodes 110 via a fronthaul link or a midhaul link, such as a DU. In some examples, the network node 110 is or includes a network node that communicates with other network nodes 110 via a midhaul link or communicates with the core network node 160 via a backhaul link, such as a CU. In some examples, the network node 110 (such as an aggregated network node 110 or a decomposed network node 110) may include multiple network nodes, such as one or more RUs, one or more CUs, and / or one or more DUs. For example, the network node 110 may include an NR base station, an LTE base station, a Node B, an eNB (e.g., in 4G), a gNB (e.g., in 5G), an access point, a transmit receive point (TRP), a DU, a RU, a CU, a mobility element of a network, a core network node 160, a network element, a network equipment, a RAN node, or a combination thereof. In some examples, network nodes 110 may be interconnected to each other or to one or more other network nodes 110 in wireless network 100 via various types of fronthaul, midhaul, and / or backhaul interfaces, such as direct physical connections, air interfaces, or virtual networks, using any suitable transport network.

[0047] In some examples, the network node 110 may provide communication coverage for a particular geographic area. In the Third Generation Partnership Project (3GPP), the term "cell" may refer to the coverage area of ​​the network node 110 and / or the network node subsystem serving the coverage area, depending on the context in which the term is used. The network node 110 may provide communication coverage for a macro cell, a pico cell, a femto cell, and / or another type of cell. A macro cell may cover a relatively large geographic area (e.g., a radius of several kilometers) and may allow unrestricted access by a UE 120 with a service subscription. A pico cell may cover a relatively small geographic area and may allow unrestricted access by a UE 120 with a service subscription. A femto cell may cover a relatively small geographic area (e.g., a home) and may allow restricted access by a UE 120 associated with the femto cell (e.g., a UE 120 in a closed subscriber group (CSG)). A network node 110 for a macro cell may be referred to as a macro network node. A network node 110 for a pico cell may be referred to as a pico network node. The network node 110 for a femto cell may be referred to as a femto network node or a home network node. Figure 1 In the example shown in , network node 110a may be a macro network node for macro cell 102a, network node 110b may be a pico network node for pico cell 102b, and network node 110c may be a femto network node for femto cell 102c. A network node may support one or more (e.g., three) cells. In some examples, a cell may not necessarily be stationary, and the geographic area of ​​a cell may move depending on the location of a mobile network node 110 (e.g., a mobile network node).

[0048] In some aspects, the term "base station" or "network node" may refer to an aggregated base station, a decomposed base station, an integrated access and backhaul (IAB) node, a relay node, or one or more components thereof. For example, in some aspects, a "base station" or "network node" may refer to a CU, a DU, a RU, a near real-time (near RT) RAN intelligent controller (RIC), or a non-real-time (non-RT) RIC, or a combination thereof. In some aspects, the term "base station" or "network node" may refer to a device configured to perform one or more functions (such as those described herein in conjunction with the network node 110). In some aspects, the term "base station" or "network node" may refer to a plurality of devices configured to perform one or more functions. For example, in some distributed systems, each of a number of different devices (which may be located in the same geographic location or in different geographic locations) may be configured to perform at least a portion of a function, or to repeat the execution of at least a portion of the function, and the term "base station" or "network node" may refer to any one or more of these different devices. In some aspects, the term "base station" or "network node" may refer to one or more virtual base stations or one or more virtual base station functions. For example, in some aspects, two or more base station functions may be instantiated on a single device. In some aspects, the term "base station" or "network node" may refer to one of the base station functions but not the other. In this way, a single device may include more than one base station.

[0049] The wireless network 100 may include one or more relay stations. A relay station is a network node that can receive transmissions of data from an upstream node (e.g., a network node 110 or a UE 120) and transmit transmissions of data to a downstream node (e.g., a UE 120 or a network node 110). A relay station may be a UE 120 that can relay transmissions for other UEs 120. Figure 1 In the example shown in , a network node 110d (e.g., a relay network node) may communicate with a network node 110a (e.g., a macro network node) and a UE 120d to facilitate communication between the network node 110a and the UE 120d. A network node 110 that relays communications may be referred to as a relay station, a relay base station, a relay network node, a relay node, a relay, etc.

[0050] The wireless network 100 may be a heterogeneous network that includes different types of network nodes 110, such as macro network nodes, pico network nodes, femto network nodes, relay network nodes, etc. These different types of network nodes 110 may have different transmit power levels, different coverage areas, and / or different impacts on interference in the wireless network 100. For example, a macro network node may have a high transmit power level (e.g., 5 watts to 40 watts), while a pico network node, a femto network node, and a relay network node may have a lower transmit power level (e.g., 0.1 watt to 2 watts).

[0051] The network controller 130 may be coupled to or in communication with a set of network nodes 110 and may provide coordination and control for the network nodes 110. The network controller 130 may communicate with the network nodes 110 via a backhaul communication link or a midhaul communication link. The network nodes 110 may also communicate directly with each other or indirectly via a wireless backhaul communication link or a wired backhaul communication link. In some aspects, the network controller 130 may be or may include a CU or a core network node 160.

[0052] UE 120 can be distributed throughout the wireless network 100, and each UE 120 can be stationary or mobile. UE 120 can include, for example, an access terminal, a terminal, a mobile station and / or a subscriber unit. UE 120 can be a cellular phone (e.g., a smart phone), a personal digital assistant (PDA), a wireless modem, a wireless communication device, a handheld device, a laptop computer, a cordless phone, a wireless local loop (WLL) station, a tablet computer, a camera, a gaming device, a netbook, a smartbook, an ultrabook, a medical device, a biometric device, a wearable device (e.g., a smart watch, smart clothing, smart glasses, a smart wristband, smart jewelry (e.g., a smart ring or a smart bracelet)), an entertainment device (e.g., a music device, a video device and / or a satellite radio), a vehicle component or sensor, a smart meter / sensor, an industrial manufacturing equipment, a global positioning system device, a UE function of a network node and / or any other suitable device configured to communicate via a wireless or wired medium.

[0053] Some UEs 120 may be considered as machine type communication (MTC) or evolved or enhanced machine type communication (eMTC) UEs. MTC UEs and / or eMTC UEs may include, for example, robots, drones, remote devices, sensors, meters, monitors, and / or location tags that may communicate with a network node, another device (e.g., a remote device), or some other entity. Some UEs 120 may be considered as Internet of Things (IoT) devices and / or may be implemented as NB-IoT (narrowband IoT) devices. Some UEs 120 may be considered as customer premises equipment. UE 120 may be included inside a housing that houses components of UE 120, such as a processor component and / or a memory component. In some examples, the processor component and the memory component may be coupled together. For example, a processor component (e.g., one or more processors) and a memory component (e.g., a memory) may be operably coupled, communicatively coupled, electronically coupled, and / or electrically coupled.

[0054] In general, any number of wireless networks 100 may be deployed in a given geographic area. Each wireless network 100 may support a specific RAT and may operate on one or more frequencies. RAT may be referred to as a radio technology, air interface, etc. Frequency may be referred to as a carrier, frequency channel, etc. Each frequency may support a single RAT in a given geographic area to avoid interference between wireless networks of different RATs. In some cases, NR or 5G RAT networks may be deployed.

[0055] In some examples, two or more UEs 120 (e.g., shown as UE 120a and UE 120e) may communicate directly (e.g., without using network node 110 as an intermediary to communicate with each other) using one or more side link channels. For example, UE 120 may communicate using peer-to-peer (P2P) communication, device-to-device (D2D) communication, vehicle-to-everything (V2X) protocols (e.g., which may include vehicle-to-vehicle (V2V) protocols, vehicle-to-infrastructure (V2I) protocols, or vehicle-to-pedestrian (V2P) protocols), and / or mesh networks. In such examples, UE 120 may perform scheduling operations, resource selection operations, and / or other operations described elsewhere herein as being performed by network node 110.

[0056] The devices of the wireless network 100 may communicate using an electromagnetic spectrum that may be subdivided into various categories, bands, channels, etc., based on frequency or wavelength. For example, the devices of the wireless network 100 may communicate using one or more operating bands. In 5G NR, two initial operating bands have been identified as frequency ranges designated FR1 (410 MHz–7.125 GHz) and FR2 (24.25 GHz–52.6 GHz). It should be understood that, although a portion of FR1 is greater than 6 GHz, FR1 is often (interchangeably) referred to as the “below 6 GHz” band in various documents and articles. A similar naming problem sometimes occurs with respect to FR2, which is often (interchangeably) referred to as the “millimeter wave” band in documents and articles, although different from the extremely high frequency (EHF) band (30 GHz–300 GHz) identified as the “millimeter wave” band by the International Telecommunication Union (ITU).

[0057] Frequencies between FR1 and FR2 are generally referred to as mid-band frequencies. Recent 5G NR research has identified the operating bands for these mid-band frequencies as frequency range designation FR3 (7.125GHz–24.25GHz). The bands falling within FR3 can inherit FR1 characteristics and / or FR2 characteristics, so the features of FR1 and / or FR2 can be effectively extended to mid-band frequencies. In addition, higher frequency bands are currently being explored to extend 5G NR operations to more than 52.6GHz. For example, three higher operating bands have been identified as frequency range designations FR4a or FR4-1 (52.6GHz–71GHz), FR4 (52.6GHz–114.25GHz), and FR5 (114.25GHz–300GHz). Each of these higher frequency bands falls within the EHF band.

[0058] Considering the above examples, unless otherwise specifically stated, it should be understood that if the term "below 6 GHz" or the like is used herein, the term may broadly refer to frequencies that may be lower than 6 GHz, may be within FR1, or may include mid-band frequencies. In addition, unless otherwise specifically stated, it should be understood that if the term "millimeter wave" or the like is used herein, the term may broadly refer to frequencies that may include mid-band frequencies, may be within FR2, FR4, FR4-a or FR4-1, and / or FR5, or may be within the EHF band. It is contemplated that the frequencies included in these operating bands (e.g., FR1, FR2, FR3, FR4, FR4-a, FR4-1, and / or FR5) may be modified, and the techniques described herein are applicable to those modified frequency ranges.

[0059] In some aspects, UE 120 may include a communications manager 140. As described in greater detail elsewhere herein, communications manager 140 may send first machine learning data associated with a machine learning service to a first network node for use by a first inference host associated with the first network node; receive a handover command communication from the first network node indicating that the UE is to perform a handover from the first network node to a second network node, wherein the handover command communication indicates machine learning inference information associated with a second inference host associated with the second network node; and send second machine learning data associated with the machine learning service to the second network node for use by the second inference host associated with the second network node based at least in part on receiving the handover command communication. Additionally or alternatively, communications manager 140 may perform one or more other operations described herein.

[0060] In some aspects, the first network node 110 may include a communications manager 150. As described in greater detail elsewhere herein, the communications manager 150 may receive a first handover communication from a first inference host associated with the first network node, the first handover communication indicating first machine learning inference information associated with a machine learning service associated with machine learning data associated with a first UE; send a handover request communication associated with a handover procedure for the first UE to a second network node or a core network node, wherein the first network node selects the second network node for the handover procedure based at least in part on the first machine learning inference information and the machine learning service associated with the second network node; and receive a handover response communication associated with the handover procedure from the second network node or the core network node, the handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node. As another example, the communication manager 150 may receive a handover request communication associated with a handover process for a UE from a second network node or a core network node, wherein the handover request communication indicates first machine learning inference information associated with a machine learning service and machine learning data associated with the UE; communicate with a first inference host associated with the first network node to determine that the first inference host can support the machine learning service; and send a handover response communication indicating second machine learning inference information associated with the first inference host to the second network node or the core network node based at least in part on the first inference host being able to support the machine learning service. Additionally or alternatively, the communication manager 150 may perform one or more other operations described herein.

[0061] In some aspects, the core network node 160 may include a communication manager 170. As described in more detail elsewhere herein, the communication manager 170 may receive a first handover request communication associated with a handover procedure for a UE from a first network node, wherein the first handover request communication indicates at least one of first machine learning inference information associated with a machine learning service, machine learning data associated with the UE, or a first inference host associated with the first network node; send a second handover request communication associated with the handover procedure for the UE to a second network node, wherein the second network node is selected for the handover procedure based at least in part on the first machine learning inference information and a machine learning capability associated with the second network node; and receive a first handover response communication associated with the handover procedure from the second network node, the first handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node. Additionally or alternatively, the communication manager 170 may perform one or more other operations described herein.

[0062] As indicated above, Figure 1 are provided as examples. Other examples can be found in the Figure 1 Different than described.

[0063] Figure 2 2 is a diagram illustrating an example 200 of a network node 110 communicating with a UE 120 in a wireless network 100 according to the present disclosure. The network node 110 may be equipped with a set of antennas 234a to 234t, such as T antennas (T≥1). The UE 120 may be equipped with a set of antennas 252a to 252r, such as R antennas (R≥1). The network node 110 of example 200 includes one or more radio frequency components, such as an antenna 234 and a modem 232. In some examples, the network node 110 may include an interface, a communication component, or another component that facilitates communication with the UE 120 or another network node. Some network nodes 110 may not include a radio frequency component that facilitates direct communication with the UE 120, such as one or more CUs or one or more DUs.

[0064] At the network node 110, a transmit processor 220 may receive data intended for a UE 120 (or a set of UEs 120) from a data source 212. The transmit processor 220 may select one or more modulation and coding schemes (MCS) for the UE 120 based at least in part on one or more channel quality indicators (CQIs) received from the UE 120. The network node 110 may process (e.g., encode and modulate) the data for the UE 120 based at least in part on the MCS selected for the UE 120, and provide data symbols for the UE 120. The transmit processor 220 may process system information (e.g., for semi-static resource partitioning information (SRPI)) and control information (e.g., CQI requests, grants, and / or upper layer signaling), and provide overhead symbols and control symbols. The transmit processor 220 may generate reference symbols for reference signals (e.g., cell-specific reference signals (CRS) or demodulation reference signals (DMRS)) and synchronization signals (e.g., primary synchronization signals (PSS) or secondary synchronization signals (SSS)). The transmit (TX) multiple-input multiple-output (MIMO) processor 230 may perform spatial processing (e.g., precoding) on ​​data symbols, control symbols, overhead symbols, and / or reference symbols, where applicable, and may provide a set of output symbol streams (e.g., T output symbol streams) to a corresponding set of modems 232 (e.g., T modulators) (shown as modems 232a to 232t). For example, each output symbol stream may be provided to a modulator component (shown as MOD) of the modem 232. Each modem 232 may process a corresponding output symbol stream (e.g., for OFDM) using a corresponding modulator component to obtain an output sample stream. Each modem 232 may also process (e.g., convert to analog, amplify, filter, and / or upconvert) the output sample stream using a corresponding modulator component to obtain a downlink signal. The modems 232a through 232t may transmit a set of downlink signals (eg, T downlink signals) via a corresponding set of antennas 234 (eg, T antennas) (shown as antennas 234a through 234t).

[0065] At the UE 120, a set of antennas 252 (shown as antennas 252a to 252r) may receive downlink signals from the network node 110 and / or other network nodes 110 and may provide a set of received signals (e.g., R received signals) to a set of modems 254 (e.g., R modems) (shown as modems 254a to 254r). For example, each received signal may be provided to a demodulator component (shown as DEMOD) of the modem 254. Each modem 254 may use a corresponding demodulator component to condition (e.g., filter, amplify, downconvert, and / or digitize) the received signal to obtain input samples. Each modem 254 may use a demodulator component to further process the input samples (e.g., for OFDM) to obtain received symbols. A MIMO detector 256 may obtain received symbols from the modem 254, may perform MIMO detection on the received symbols where applicable, and may provide detected symbols. The receive processor 258 may process (e.g., demodulate and decode) the detected symbols, may provide decoded data for the UE 120 to the data sink 260, and may provide decoded control information and system information to the controller / processor 280. The term "controller / processor" may refer to one or more controllers, one or more processors, or a combination thereof. The channel processor may determine a reference signal received power (RSRP) parameter, a received signal strength indicator (RSSI) parameter, a reference signal received quality (RSRQ) parameter, and / or a CQI parameter, among other things. In some examples, one or more components of the UE 120 may be included in the housing 284.

[0066] The network controller 130 may include a communication unit 294, a controller / processor 290, and a memory 292. The network controller 130 may include, for example, one or more devices in a core network. The network controller 130 may communicate with the network node 110 via the communication unit 294.

[0067] One or more antennas (e.g., antennas 234a to 234t and / or antennas 252a to 252r) may include or be included within one or more antenna panels, one or more antenna groups, one or more sets of antenna elements, and / or one or more antenna arrays, etc. Antenna panels, antenna groups, sets of antenna elements, and / or antenna arrays may include one or more antenna elements (in a single housing or multiple housings), sets of coplanar antenna elements, sets of non-coplanar antenna elements, and / or may be coupled to one or more transmit and / or receive components (such as, Figure 2 One or more antenna elements of one or more components in.

[0068] On the uplink, at the UE 120, the transmit processor 264 may receive and process data from the data source 262 and control information from the controller / processor 280 (e.g., for reports including RSRP, RSSI, RSRQ, and / or CQI). The transmit processor 264 may generate reference symbols for one or more reference signals. The symbols from the transmit processor 264 may be pre-decoded by the TX MIMO processor 266, where applicable, further processed by the modem 254 (e.g., for DFT-s-OFDM or CP-OFDM), and sent to the network node 110. In some examples, the modem 254 of the UE 120 may include a modulator and a demodulator. In some examples, the UE 120 includes a transceiver. The transceiver may include any combination of an antenna 252, a modem 254, a MIMO detector 256, a receive processor 258, a transmit processor 264, and / or a TX MIMO processor 266. The transceiver may be used by a processor (eg, controller / processor 280) and memory 282 to perform the functions described herein (eg, reference Figure 8 , Fig. 9A , Fig. 9B as well as Figures 10 to 17 ) or any aspects of any of the methods described herein.

[0069] At the network node 110, uplink signals from the UE 120 and / or other UEs may be received by the antenna 234, processed by the modem 232 (e.g., a demodulator component of the modem 232 shown as DEMOD), detected by the MIMO detector 236 (where applicable), and further processed by the receive processor 238 to obtain decoded data and control information transmitted by the UE 120. The receive processor 238 may provide the decoded data to the data sink 239 and the decoded control information to the controller / processor 240. The network node 110 may include a communication unit 244 and may communicate with the network controller 130 via the communication unit 244. The network node 110 may include a scheduler 246 to schedule one or more UEs 120 for downlink and / or uplink communications. In some examples, the modem 232 of the network node 110 may include a modulator and a demodulator. In some examples, the network node 110 includes a transceiver. The transceiver may include any combination of antenna 234, modem 232, MIMO detector 236, receive processor 238, transmit processor 220, and / or TX MIMO processor 230. The transceiver may be used by a processor (e.g., controller / processor 240) and memory 242 to execute the instructions described herein (e.g., reference 2 Figure 8 , Fig. 9A , Fig. 9B as well as Figures 10 to 17 ) or any aspects of any of the methods described herein.

[0070] The controller / processor 240 of the network node 110, the controller / processor 280 of the UE 120, and / or Figure 2 Any other components of the network node 110 may perform one or more techniques associated with UE machine learning service continuity, as described in more detail elsewhere herein. For example, the controller / processor 240 of the network node 110, the controller / processor 280 of the UE 120, and / or Figure 2 Any other component of the Fig.11 Process 1100, Fig.12 The process 1200 Fig.13 The process of 1300 Fig.14 1400 and / or operations of other processes as described herein. Memory 242 and memory 282 may store data and program codes for network node 110 and UE 120, respectively. In some examples, memory 242 and / or memory 282 may include a non-transitory computer-readable medium storing one or more instructions (e.g., code and / or program code) for wireless communication. For example, the one or more instructions, when executed (e.g., directly, or after compilation, conversion, and / or interpretation) by one or more processors of network node 110 and / or UE 120, may cause the one or more processors, UE 120, and / or network node 110 to perform or direct, for example, Fig.11 Process 1100, Fig.12 The process 1200 Fig.13 Process 1300, Fig.14 The process 1400 and / or operations of other processes as described herein. In some examples, executing instructions may include running instructions, converting instructions, compiling instructions, and / or interpreting instructions, etc.

[0071] In some aspects, UE 120 includes means for sending first machine learning data associated with a machine learning service to a first network node for use by a first inference host associated with the first network node; means for receiving a handover command communication from the first network node indicating that the UE will perform a handover from the first network node to a second network node, wherein the handover command communication indicates machine learning inference information associated with a second inference host associated with the second network node; and / or means for sending second machine learning data associated with the machine learning service to the second network node for use by the second inference host associated with the second network node based at least in part on receiving the handover command communication. Means for UE 120 to perform operations described herein may include, for example, one or more of communications manager 140, antenna 252, modem 254, MIMO detector 256, receive processor 258, transmit processor 264, TX MIMO processor 266, controller / processor 280, or memory 282.

[0072] In some aspects, the first network node 110 includes a component for receiving a first handover communication from a first inference host associated with the first network node, the first handover communication indicating first machine learning inference information associated with a machine learning service, the machine learning service associated with machine learning data associated with a first UE; a component for sending a handover request communication associated with a handover procedure for the first UE to a second network node or a core network node, wherein the first network node selects the second network node for the handover procedure based at least in part on the first machine learning inference information and the machine learning service associated with the second network node; and / or a component for receiving a handover response communication associated with the handover procedure from the second network node or the core network node, the handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node. In some aspects, means for the first network node 110 to perform operations described herein may include, for example, one or more of the communication manager 150, the transmit processor 220, the TX MIMO processor 230, the modem 232, the antenna 234, the MIMO detector 236, the receive processor 238, the controller / processor 240, the memory 242, or the scheduler 246.

[0073] In some aspects, the first network node 110 includes means for receiving a handover request communication associated with a handover procedure for a UE from a second network node or a core network node, wherein the handover request communication indicates first machine learning inference information associated with a machine learning service and machine learning data associated with the UE; means for communicating with a first inference host associated with the first network node to determine that the first inference host is capable of supporting the machine learning service; and / or means for sending a handover response communication indicating second machine learning inference information associated with the first inference host to the second network node or the core network node based at least in part on the first inference host being capable of supporting the machine learning service. In some aspects, means for the first network node 110 to perform operations described herein may include, for example, one or more of the communication manager 150, the transmit processor 220, the TX MIMO processor 230, the modem 232, the antenna 234, the MIMO detector 236, the receive processor 238, the controller / processor 240, the memory 242, or the scheduler 246.

[0074] In some aspects, the core network node 160 includes means for receiving a first handover request communication associated with a handover procedure for a UE from a first network node, wherein the first handover request communication indicates at least one of first machine learning inference information associated with a machine learning service, machine learning data associated with the UE, or a first inference host associated with the first network node; means for sending a second handover request communication associated with the handover procedure for the UE to a second network node, wherein the second network node is selected for the handover procedure based at least in part on the first machine learning inference information and a machine learning capability associated with the second network node; and / or means for receiving a first handover response communication associated with the handover procedure from the second network node, the first handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node. In some aspects, the means for the core network node to perform operations described herein may include, for example, one or more of the communication manager 170, a transmit processor, a TX MIMO processor, a modem, an antenna, a MIMO detector, a receive processor, a controller / processor, a memory, or a scheduler.

[0075] Although Figure 2 The blocks in the 2000 and 2010 are illustrated as distinct components, but the functionality described above for these blocks may be implemented in a single hardware, software, or combined component or in various combinations of components. For example, the functionality described for the transmit processor 264, the receive processor 258, and / or the TX MIMO processor 266 may be performed by or under the control of the controller / processor 280.

[0076] As indicated above, Figure 2 are provided as examples. Other examples can be found in the Figure 2 Different than described.

[0077] The deployment of a communication system (such as a 5G NR system) can be arranged with various components or components in a variety of ways. In a 5G NR system or network, a network node, a network entity, a mobility element of a network, a RAN node, a core network node, a network element, a base station or network equipment can be implemented in an aggregated or decomposed architecture. For example, a base station (such as a node B (NB), an evolved NB (eNB), an NR BS, a 5G NB, an access point (AP), a TRP or a cell, etc.) or one or more units (or one or more components) that perform base station functionality can be implemented as an aggregated base station (also referred to as an independent base station or a monolithic base station) or a decomposed base station. "Network entity" or "network node" may refer to a decomposed base station or one or more units of a decomposed base station (such as one or more CUs, one or more DUs, one or more RUs, or a combination thereof).

[0078] An aggregated base station (e.g., an aggregated network node) may be configured to utilize a radio protocol stack physically or logically integrated within a single RAN node (e.g., within a single device or unit). A decomposed base station (e.g., a decomposed network node) may be configured to utilize a protocol stack physically or logically distributed between two or more units (such as one or more CUs, one or more DUs, or one or more RUs). In some examples, a CU may be implemented within a network node, and one or more DUs may be co-located with the CU, or alternatively, may be geographically or virtually distributed across one or more other network nodes. A DU may be implemented to communicate with one or more RUs. Each of a CU, a DU, and a RU may also be implemented as a virtual unit, such as a virtual central unit (VCU), a virtual distributed unit (VDU), or a virtual radio unit (VRU), and the like.

[0079] Base station type operations or network designs may take into account the aggregated nature of base station functionality. For example, a decomposed base station may be utilized in an IAB network, an open radio access network (O-RAN (such as a network configuration initiated by the O-RAN Alliance)), or a virtualized radio access network (vRAN, also known as a cloud radio access network (C-RAN)) to facilitate the scaling of a communication system by separating base station functionality into one or more units that can be deployed separately. A decomposed base station may include functionality implemented across two or more units at various physical locations, as well as functionality implemented virtually for at least one unit, which may enable flexibility in network design. Individual units of a decomposed base station may be configured for wired or wireless communication with at least one other unit of the decomposed base station.

[0080] Figure 3 3 is a diagram illustrating an example disaggregated base station architecture 300 according to the present disclosure. The disaggregated base station architecture 300 may include a CU 310, which may communicate directly with a core network 320 via a backhaul link, or indirectly with the core network 320 through one or more disaggregated control units, such as a near real-time (RT) RAN intelligent controller (RIC) 325 via an E2 link, or a non-real-time (RT) RIC 315 associated with a service management and orchestration (SMO) framework 305, or both. The CU 310 may communicate with one or more DUs 330 via respective midhaul links, such as via an F1 interface. Each of the DUs 330 may communicate with one or more RUs 340 via respective fronthaul links. Each of the RUs 340 may communicate with one or more UEs 120 via respective radio frequency (RF) access links. In some implementations, a UE 120 may be served simultaneously by multiple RUs 340.

[0081] Each of the units (including CU 310, DU 330, RU 340) and the near-RT RIC 325, non-RT RIC 315, and SMO framework 305 may include or be coupled to one or more interfaces configured to receive or send signals, data, or information (collectively referred to as signals) via a wired or wireless transmission medium. Each of the units or an associated processor or controller that provides instructions to one or more communication interfaces of the corresponding unit may be configured to communicate with one or more of the other units via a transmission medium. In some examples, each of these units may include a wired interface and a wireless interface, the wired interface being configured to receive signals or send signals to one or more of the other units via a wired transmission medium, and the wireless interface may include a receiver, a transmitter, or a transceiver (such as an RF transceiver), which is configured to receive signals or send signals to one or more of the other units via a wireless transmission medium, or both.

[0082] In some aspects, CU 310 may host one or more higher layer control functions. Such control functions may include radio resource control (RRC) functions, packet data convergence protocol (PDCP) functions, or service data adaptation protocol (SDAP) functions, etc. Each control function may be implemented using an interface that is configured to communicate signals with other control functions hosted by CU 310. CU 310 may be configured to handle user plane functionality (e.g., central unit-user plane (CU-UP) functionality), control plane functionality (e.g., central unit-control plane (CU-CP) functionality), or a combination thereof. In some specific implementations, CU 310 may be logically split into one or more CU-UP units and one or more CU-CP units. When implemented in an O-RAN configuration, the CU-UP unit may communicate bidirectionally with the CU-CP unit via an interface such as an E1 interface. As needed, CU 310 may be implemented to communicate with DU 330 for network control and signaling.

[0083] Each DU 330 may correspond to a logical unit that includes one or more base station functions for controlling the operation of one or more RUs 340. In some aspects, the DU 330 may host one or more of a radio link control (RLC) layer, a MAC layer, and one or more high physical (PHY) layers, at least in part, according to a functional partition such as that defined by 3GPP. In some aspects, one or more high PHY layers may be implemented by one or more modules for forward error correction (FEC) encoding and decoding, scrambling, modulation and demodulation, etc. In some aspects, the DU 330 may also host one or more low PHY layers, such as one or more modules for fast Fourier transform (FFT), inverse FFT (iFFT), digital beamforming or physical random access channel (PRACH) extraction and filtering, etc. Each layer (which may also be referred to as a module) may be implemented using an interface that is configured to communicate signals with other layers (and modules) hosted by the DU 330 or with control functions hosted by the CU 310.

[0084] Each RU 340 may implement lower layer functionality. In some deployments, the RU 340 controlled by the DU 330 may correspond to a logical node that hosts RF processing functions based on functional splits (e.g., functional splits defined by 3GPP) such as lower layer functional splits, or low PHY layer functions such as performing FFT, performing iFFT, digital beamforming, or PRACH extraction and filtering, etc. In this architecture, each RU 340 may be operated to handle over-the-air (OTA) communications with one or more UEs 120. In some specific implementations, real-time and non-real-time aspects of control plane and user plane communications with the RU 340 may be controlled by the corresponding DU 330. In some scenarios, this configuration may enable each DU 330 and CU 310 to be implemented in a cloud-based RAN architecture such as a vRAN architecture.

[0085] The SMO framework 305 may be configured to support RAN deployment and provisioning of non-virtualized network elements and virtualized network elements. For non-virtualized network elements, the SMO framework 305 may be configured to support the deployment of dedicated physical resources for RAN coverage requirements, which may be managed via an operation and maintenance interface (such as an O1 interface). For virtualized network elements, the SMO framework 305 may be configured to interact with a cloud computing platform (such as an open cloud (O-Cloud) platform 390) to perform network element lifecycle management (such as instantiating virtualized network elements) via a cloud computing platform interface (such as an O2 interface). Such virtualized network elements may include, but are not limited to, CU 310, DU 330, RU 340, non-RT RIC 315, and near-RT RIC 325. In some specific implementations, the SMO framework 305 may communicate with hardware aspects of the 4G RAN (such as an open eNB (O-eNB) 311) via the O1 interface. Additionally, in some implementations, the SMO framework 305 can communicate directly with each of the one or more RUs 340 via a corresponding O1 interface. The SMO framework 305 can also include a non-RT RIC 315 configured to support the functionality of the SMO framework 305. In some examples, the SMO framework 305 can be connected (e.g., via an O1* interface) to an infrastructure management framework that supports O-RAN virtual network functions.

[0086] The non-RT RIC 315 may be configured to include logic functions that implement non-real-time control and optimization of RAN elements and resources, artificial intelligence / machine learning (AI / ML) workflows including model training and updating, or policy-based guidance of applications / features in the near-RT RIC 325. The non-RT RIC 315 may be coupled to or communicate with the near-RT RIC 325 (such as via an A1 interface). The near-RT RIC 325 may be configured to include logic functions that implement near-real-time control and optimization of RAN elements and resources via data collection and actions through an interface (such as via an E2 interface) that connects one or more CUs 310, one or more DUs 330, or both, and the O-eNB with the near-RT RIC 325. In some examples, the near-RT RIC 325 may be a logic function that implements near-real-time control and optimization of O-RAN elements and resources via fine-grained data collection and actions through the E2 interface. The near-RT RIC 325 may be co-located with the RAN or network entity to provide real-time processing, such as online ML training or near real-time ML inference. The non-RT RIC 315 may be a logic function that implements non-real-time control and optimization of RAN elements and resources, policy-based guidance of AI / ML workflows including model training and updates and applications / features in the near-RT RIC 325, and ML inference with less latency specifications. The non-RT RIC 315 may be located farther from the RAN or base station, such as on a cloud-based server or on an edge server.

[0087] In some implementations, in order to generate an AI / ML model to be deployed in the near-RT RIC 325, the non-RT RIC 315 may receive parameters or external enrichment information from an external server. Such information may be utilized by the near-RT RIC 325 and may be received from a non-network data source or from a network function at the SMO framework 305 or the non-RT RIC 315. In some examples, the non-RT RIC 315 or the near-RT RIC 325 may be configured to tune RAN behavior or performance. For example, the non-RT RIC 315 may monitor long-term trends and patterns of performance and employ AI / ML models to perform corrective actions through the SMO framework 305 (such as via reconfiguration of the O1 interface) or via the creation of RAN management policies (such as A1 interface policies).

[0088] In some examples, the overall RIC architecture may consist of four functional software elements: DU software functionality, multi-RAT CU protocol stack, near-RT RIC 325 itself, and an orchestration / network management system (NMS) layer with non-RT RIC 315. The functional software elements may interact with the RU hardware to enable it to run more efficiently and be optimized in real time as part of the RAN cluster to deliver a better network experience to end users.

[0089] In some examples, the functionality of the non-RT RIC 315 may include configuration management, device management, fault management, performance management and lifecycle management for all network elements in the network. The non-RT RIC 315 may use data analysis and AI / ML training / inference to determine RAN optimization actions, for which it may utilize services of the SMO framework 305, such as data collection and provisioning services for network nodes. The near-RT RIC 325 may use embedded intelligence for load balancing, resource block (RB) management, interference detection and / or interference mitigation per UE control, and the like. The near-RT RIC 325 may also provide quality of service (QoS) management, connection management and / or seamless handover control, and the like. The near-RT RIC 325 may utilize the near real-time status of the underlying network and feed RAN data to train an AI / ML model, which may then be fed to the near-RT RIC 325 to facilitate radio resource management for subscribers.

[0090] As indicated above, Figure 3 are provided as examples. Other examples can be found in the Figure 3 Different than described.

[0091] Wireless networks can operate in higher frequency bands, such as in millimeter wave (mmW) bands (e.g., FR2 above 28 GHz, FR4 above 60 GHz, or THz bands above 100 GHz, etc.), to provide high data rates. For example, wireless devices (such as network nodes and UEs) can communicate with each other through beamforming technology to improve communication speed and reliability. Beamforming technology enables wireless devices to send signals in a specific direction instead of sending omnidirectional signals in all directions. In some examples, a wireless device can use a common wavelength and phase for transmitting from multiple antenna elements to send signals from the multiple antenna elements, and can combine the signals from the multiple antenna elements to produce a combined signal with a longer range and a more directional beam. The beam width of the signal can vary based on the transmission frequency. For example, the width of the beam can be negatively correlated with the frequency, where the beam width can decrease as the transmission frequency increases because more radiating elements can be placed in each given area at the transmitter due to smaller wavelengths. Thus, higher frequency bands (e.g., THz or sub-THz bands) may enable wireless devices to form much narrower beam structures (e.g., pencil beams, laser beams, or narrow beams, etc.) than beam structures at or below FR2 because more radiating elements may be placed per given area at the antenna element due to the smaller wavelength. Higher frequency bands may have short delay spreads (e.g., a few nanoseconds) and may translate into coherent frequency bandwidths of tens (10) MHz. In addition, higher frequency bands may provide large available bandwidths that may be occupied by larger bandwidth carriers, such as 1000 MHz or more per carrier. In some examples, the transmission path of a narrower beam is more likely to be customized for a receiver, making it more likely that the transmission will meet line-of-sight (LOS) conditions because a narrower beam is more likely to reach the receiver without being blocked by obstacles. Moreover, since the transmission path may be narrow, reflections and / or refractions are less likely to occur for a narrower beam.

[0092] Although higher frequency bands can provide narrower beam structures and higher transmission rates, higher frequency bands may also encounter higher attenuation and diffraction losses, where the blocking of the LOS path may degrade the quality of the wireless link. For example, when two wireless devices communicate with each other based on the LOS path in a higher frequency band and the LOS path is blocked by obstacles such as pedestrians, buildings and / or vehicles, the received power may drop significantly. Therefore, compared with lower frequency bands, wireless communications based on higher frequency bands may be more susceptible to environmental changes.

[0093] The various aspects presented herein can improve the performance and reliability of wireless communications operating in higher frequency bands (such as millimeter wave bands). The various aspects presented herein can enable wireless communications to adapt to dynamic environments, where wireless devices can manage wireless communications based at least in part on environmental conditions, such as performing beam management. For example, in one aspect of the present disclosure, in order to overcome the rapid changes in link quality of a wireless communication system operating in a higher frequency band caused by LOS path obstruction, sensing information provided by an onboard sensor of a vehicle (e.g., a camera, radar, and / or light detection and ranging (lidar) sensor) can be used to provide information associated with the communication environment and obstacles (e.g., moving objects) that may potentially block the LOS path and degrade the communication quality of a wireless device (such as a UE). The vehicle may include a motor vehicle (e.g., a motorcycle, a car, a truck, and / or a bus), a rail vehicle (a train, a tram, and / or a subway vehicle, etc.), a water vehicle (a ship and / or a boat), and / or an aircraft (e.g., an airplane, a helicopter, an air floater, and / or a drone), etc. In addition, in another aspect of the present disclosure, the vehicle may be configured to employ onboard computing resources and machine learning (ML) models to pre-process the collected sensor data, and feed the inference data to an inference host co-located / associated with a network node for predicting / estimating possible blocking (e.g., LOS path blocking) or optimal beam pairs so that the base station can proactively initiate beam management or handover procedures for the UE. For purposes of the present disclosure, "inference" or "ML inference" may refer to the process of running data points into an ML model (e.g., via an inference host) to compute an output such as a single numeric score (e.g., using a trained ML algorithm to make a prediction). "Inference host" or "ML inference host" may refer to a network function that hosts an ML model during inference mode. Additionally or alternatively, in another aspect of the present disclosure, a vehicle (e.g., a vehicle UE) may be configured to employ onboard computing resources and an ML model to pre-process the collected sensor data, and feed the training data to an ML training host co-located / associated with a network node for offline and / or online training of the ML model to predict / estimate possible blockages (e.g., LOS path blockages) or optimal beam pairs. For purposes of the present disclosure, "training" or "ML training" may refer to the process of running data points to train or teach an ML model (e.g., via a training host). "Training host" or "ML training host" may refer to a network function that hosts an ML model during a training mode.

[0094] Figure 4is a diagram of an example architecture 400 illustrating a functional framework for RAN intelligence implemented through data collection according to the present disclosure. In some scenarios, the functional framework for RAN intelligence may be implemented by further enhancing data collection through use cases and / or examples. For example, principles or algorithms for RAN intelligence implemented by AI / ML and associated functional frameworks (e.g., AI functionality and / or input / output of components for supporting AI optimization) have been utilized or studied to identify the benefits of an AI-enabled RAN through possible use cases (e.g., energy saving, load balancing, mobility management and / or coverage optimization, etc.). In one example, as shown in architecture 400, the functional framework for RAN intelligence may include multiple logical entities, such as a model training host 402, a model inference host 404, a data source 406, and an action party 408.

[0095] The model inference host 404 may be configured to run an AI / ML model based on the inference data provided by the data source 406, and the model inference host 404 may generate an output (e.g., a prediction) with the inference data input to the action party 408. The action party 408 may be an element or entity of the core network or RAN. For example, the action party 408 may be a base station (e.g., a gNB), a CU, a DU, and / or a RU, etc. In addition, the action party 408 may also depend on the type of task performed by the model inference host 404, the type of inference data provided to the model inference host 404, and / or the type of output generated by the model inference host 404. For example, if the output from the model inference host 404 is associated with beam management, the action party 408 may be a DU or a RU; and if the output from the model inference host 404 is associated with Tx / Rx scheduling, the action party 408 may be a CU or a DU.

[0096] After the action party 408 receives the output from the model inference host 404, the action party 408 may determine whether to take action based on the output. For example, if the action party 408 is a DU or RU and the output from the model inference host 404 is associated with beam management, the action party 408 may determine whether to change / modify the Tx / Rx beam based on the output. If the action party 408 determines to take action based on the output, the action party 408 may indicate the action to at least one action subject 410. For example, if the action party 408 determines to change / modify the Tx / Rx beam used for communication between the action party 408 and the action subject 410 (e.g., UE 120), the action party 408 may send a beam (re) configuration or beam switching indication to the action subject 410. The action party 408 may modify its Tx / Rx beam based on the beam (re) configuration, such as switching to a new Tx / Rx beam or applying different parameters to the Tx / Rx beam, etc.

[0097] The data source 406 may also be configured to collect data that is used as training data for training the ML model or as inference data for feeding the ML model inference operation. For example, the data source 406 may collect data from one or more core network and / or RAN entities that may include an action subject 410, and provide the collected data to the model training host 402 for ML model training. For example, after the action subject 410 (e.g., UE 120) receives the beam configuration from the action party 408, the action subject 410 may provide performance feedback associated with the beam configuration to the data source 406, wherein the performance feedback may be used by the model training host 402 to monitor or evaluate ML model performance, such as whether the output (e.g., prediction) provided to the action party 408 is accurate. In some examples, if the output provided by the action party 408 is inaccurate (or the accuracy is below an accuracy threshold), the model training host 402 may determine to modify or retrain the ML model used by the model inference host, such as via ML model deployment / update.

[0098] As indicated above, Figure 4 are provided as examples. Other examples can be found in the Figure 4 Different than described.

[0099] Figure 5 5 is a diagram illustrating an example 500 of improving wireless communication using data from sensors of a vehicle according to the present disclosure. A vehicle 502 equipped with a sensor device 504 (which may be referred to as a "sensor data collector") may enter a coverage area of ​​a network node 506, wherein the coverage area of ​​the network node 506 may include static objects (e.g., buildings) and non-static (e.g., mobile) objects (e.g., cars, trucks, buses, and pedestrians). The sensor device 504 may be associated with one or more sensors such as cameras, radars, ultrasonic sensors, and / or laser radars, etc. (which may be collectively referred to as "vehicle sensors"), or include the one or more sensors. In addition, the sensor device 504 may be a UE or include a UE function (e.g., a vehicle UE (VUE)), so that the sensor device 504 may communicate with a network node or network entity, such as via a Uu interface connection. As used herein, a vehicle with a UE function (such as a vehicle 502) may be referred to as a UE or a UE associated with a vehicle.

[0100] In some scenarios, the quality of the radio link between the network node 506 and the UE may be affected by both static objects and non-static objects. For example, the UE 120 in the vehicle 502 may communicate with the network node 506 based on the LOS path / link. As shown in reference numeral 510, sometimes, the communication (e.g., LOS path / link) between the UE 120 and the network node 506 may be affected by both moving objects and / or stationary objects. For example, a moving truck 512 may block the LOS path / link between the UE 120 and the network node 506, which may degrade the quality of the wireless link between the UE 120 and the network node 506. For the purposes of this disclosure, a wireless device or UE used by a user in the vehicle 502 may be referred to as a "UE in the vehicle", "end user UE" or simply "end user". A wireless device or UE placed / installed on or associated with a vehicle may be referred to as a "vehicle UE". For example, an in-vehicle UE may include a mobile phone, a tablet computer, or a wearable device, and a vehicle UE may include a vehicle onboard computer system, a vehicle sensor device / controller, or a sensor data collector, etc.

[0101] In some examples, utilizing sensory information provided by sensor devices 504 (e.g., sensor data collectors) of vehicle 502, onboard computing resources of vehicle 502, and / or ML models used by vehicle 502 to extract features, vehicle 502 may be configured to provide ML training data and / or ML inference data as well as information of UE 120 (e.g., location, speed, and / or direction of travel, etc.) to an ML training host (e.g., non-RT RIC 315, near-RT RIC 325, model training host 402, etc.) or an ML inference host 514 (e.g., near-RT RIC 325 or model inference host 404, etc.), which may be co-located or associated with network node 506 (or components of network node 506) for enhancing communication between UE 120 and network node 506, such as improving beam management for UE 120. For example, as shown by reference numeral 516, the sensor device 504 may use sensors to capture images of a view around the vehicle 502, and the captured images may be processed by an ML model 518 (e.g., an object detection model) associated with the sensor device 504 for feature extraction (e.g., for object detection).

[0102] For example, the captured image or video may be processed by an AI / ML model associated with the sensor device to generate a category map (which may also be referred to as a “segmentation map”) that identifies objects in the captured image. The category map may show that the captured image includes multiple objects, such as clouds, trees, people, aircraft, and / or buildings, among others. The category map may be further processed by an image signal processor (ISP) unit / hardware associated with the sensor device 504, wherein the ISP unit / hardware may adjust the colors of different sections of the captured image based on the category map and offline tuning data to produce a processed image (e.g., an image with extracted features). In one example of an extracted feature, a two-dimensional (2D) array of combined sensing data (e.g., detection points with (x,y) locations from a radar or lidar sensor) as an example of a point cloud (e.g., a dynamic map) from a sensor device may be processed to derive an arithmetic dynamic object, such as a bounding box (e.g., using object detection and localization based on you only look once (YOLO) from a radar point cloud).

[0103] After the sensor device 504 (e.g., a sensor data collector) extracts and processes features of one or more objects in the area, the sensor device 504 may send the extracted features as inference data to an ML inference host 514 (e.g., a near RT RIC 325 or a model inference host 404) associated with the network node 506. The ML inference host 514 may perform inferences based on the inference data using an ML model 520 (e.g., a beam prediction model) and generate output (e.g., a prediction) to the network node 506. The process of the ML inference host performing inferences based on the inference data may be referred to as providing an ML inference service or an ML service. Based at least in part on the output, the network node 506 may determine whether to perform an action. For example, the inference data from the sensor device 504 may indicate that there is a truck 512 between the UE 120 and the network node 506. Based on the inference data, the ML inference host 514 may estimate (e.g., predict) that communication between the UE 120 and the network node 506 may be improved by using a different beam pair, and the ML inference host 514 may send this estimate as an output to the network node 506. For purposes of this disclosure, a "beam pair" or a "beam pair link" may refer to a transmit beam at a transmitting device and a receive beam at a receiving device.

[0104] After receiving the output (e.g., prediction) from the ML inference host 514, the network node 506 may determine whether to take action based on the output. For example, if the network node 506 determines to take action based on the output, the network node 506 may generate a beam configuration and send the beam configuration to the UE 120. The beam configuration may indicate that the UE 120 will reconfigure the active beam or beam pair or candidate beam list and related parameters, or send a beam switching indication to the UE 120, or switch the current beam or beam pair between the UE 120 and the network node 506 to the beam or beam pair estimated / predicted by the ML inference host. Alternatively, if the network node 506 determines not to take action based on the output, the network node 506 may ignore the output and continue to communicate with the UE 120 based on the current beam or beam pair. Therefore, the network node 506 may actively manage the beam or beam pair with the UE 120 based on the beam prediction from the ML inference host 514. The beam pair link quality between UE 120 and network node 506 may be improved based on sensing information received from sensor device 504 of vehicle 502. In some examples, the sensing information received from sensor device 504 may be external data to a core network or RAN (e.g., sensor data may not be collected or derived within a communication network such as a core network or RAN).

[0105] As indicated above, Figure 5 are provided as examples. Other examples can be found in the Figure 5 Different than described.

[0106] Figure 6 600 is a diagram illustrating an example of a system architecture for utilizing sensed information from a sensor device to improve wireless communication according to various aspects of the present disclosure. Figure 5 As described, the system architecture may include at least a sensor data collector (e.g., sensor device 504) co-located with a vehicle (e.g., vehicle 502), an ML inference host (e.g., near real-time RIC 325, model inference host 404, or ML inference host 514), an actor (e.g., actor 408 or network node 506), an ML training host (e.g., non-real-time RIC 315, near real-time RIC 325, or model training host 402), and an end user (e.g., in-vehicle UE or UE 120).

[0107] In some examples, the sensor data collector may be a vehicle including one or more sensors (e.g., cameras, radars, ultrasonic sensors, and / or lidars) and an ML model for preprocessing sensor data, such as feature extraction (e.g., object detection). The sensor data collector may provide training data to an ML training host for non-real-time or offline training (e.g., an ML training host or model training host 402 at a non-real-time RIC 315) or an ML training host for online training (e.g., an ML training host or model training host 402 at a near real-time RIC 325), and the sensor data collector may provide the training data or inference data and information associated with an end user (e.g., a UE in a vehicle) to the ML training host for offline or online training or to an ML inference host for near real-time inference, respectively. The ML inference host may be co-located with an actor (e.g., a network node 506) and may make inferences (e.g., predictions) based on the inference data and information associated with the end user provided by the sensor data collector. The action party may receive an inference output (e.g., a prediction) generated by the ML inference host using an ML inference model (e.g., beam prediction), and the action party may determine whether to indicate at least one action (e.g., beam indication) to the end user. The end user (e.g., a UE in a vehicle) may take the action indicated by the action party and report the measurement to the action party accordingly. As used herein, an "ML training data service" may refer to a device (e.g., a sensor device) that provides ML training data to the ML training host, and an "ML training service" may refer to an ML training host that provides ML training to another ML function / entity (such as an ML inference host). Similarly, an "ML inference data service" may refer to a device (e.g., a sensor device) that provides ML inference data to the ML inference host, and an "ML inference service" may refer to a base station or an ML inference host that provides ML inference to one or more components, functionalities, or entities of a base station. In some examples, "ML training data service", "ML training service", "ML inference data service", and / or "ML inference service" may be collectively referred to as an "ML data service" or "machine learning service".

[0108] As indicated above, Figure 6 are provided as examples. Other examples can be found in the Figure 6 Different than described.

[0109] Figure 7is a diagram illustrating an example of a handover process 700 according to the present disclosure. The handover process 700 may be performed by UE 120, source network node 110-1, and target network node 110-2. As used herein, a "source" network node may refer to a network node with which the UE currently has an established connection (e.g., an established RRC connection). The source network node may also be referred to as a serving network node. As used herein, a "target" network node may refer to a network node that has been selected (e.g., by a source network node or a core network node) as a new source network node for the UE as part of a handover process. A "candidate" network node may refer to a network node that may be selected as a target network node for a handover process. For example, UE 120 may be configured (e.g., by source network node 110-1) to perform measurements associated with one or more candidate network nodes. UE 120 may send a measurement report indicating measurements associated with one or more candidate network nodes. In some examples, source network node 110-1 may use the measurement report to select target network node 110-2 from one or more candidate network nodes (eg, target network node 110-2 may be associated with the highest measurement value indicated by the measurement report).

[0110] like Figure 7 As shown in reference numeral 705, the source network node 110-1 may initiate a handover of the UE 120 to the target network node 110-2 by sending a handover request message to the target network node 110-2. The source network node 110-1 may send the handover request message to the target network node 110-2 via an Xn, X2, midhaul, backhaul, or next generation application protocol (NGAP) interface, etc. As shown in reference numeral 710, the target network node 110-2 may perform an admission control process associated with the handover based at least in part on receiving the handover request message. As shown in reference numeral 715, the target network node 110-2 may send a handover request confirmation message (e.g., a handover response message) to the source network node 110-1 (e.g., if the admission control process indicates that the target network node 110-2 can accept the handover of the UE 120). The handover request confirmation message may include an RRC configuration for the connection with the target network node 110-2.

[0111] As indicated by reference numeral 720, source network node 110-1 may send an RRC configuration message to UE 120 (e.g., in an RRC reconfiguration message) by forwarding the RRC configuration of the handover request confirm message to UE 120 (e.g., a handover command). As indicated by reference numeral 725, UE 120 may change the RRC connection from source network node 110-1 to target network node 110-2 based at least in part on the RRC configuration. As indicated by reference numeral 730, UE 120 may send an RRC reconfiguration complete message to target network node 110-2. The RRC reconfiguration complete message may indicate that UE 120 has changed the RRC connection from source network node 110-1 to target network node 110-2. As indicated by reference numeral 735, target network node 110-2 may send a UE context release message to source network node 110-1. The UE context release message may indicate that the handover from UE 120 to target network node 110-2 is successful.

[0112] In some examples, UE 120 may not be able to successfully connect with target network node 110-2. For example, UE 120 may attempt to connect with target network node 110-2 (e.g., by performing a random access channel (RACH) procedure with target network node 110-2), but the attempt to connect with target network node 110-2 may fail. If UE 120 is unable to successfully connect with target network node 110-2, UE 120 may perform a connection re-establishment procedure to (re)establish a connection with source network node 110-1 or another network node 110. For example, UE 120 may send an RRC re-establishment request message to the network (e.g., to source network node 110-1 or another network entity). Additionally, UE 120 may reset a medium access control (MAC) entity of UE 120, release an RRC configuration for a handover procedure, suspend all radio bearers (except SRB0 in some examples), release a connection with any configured SCell, and / or release all other configurations stored by UE 120, etc. Thus, in the event that the handover procedure with the target network node 110-2 fails, the UE 120 may (re)establish an RRC connection (eg, with the source network node 110-1 or another network node 110).

[0113] The above-described handover process 700 may ensure communication continuity for UE 120. In other words, handover process 700 may ensure that when UE 120 moves from one cell (e.g., associated with source network node 110-1) to another cell (e.g., associated with target network node 110-2) by handing over a communication connection (e.g., an RRC connection) from source network node 110-1 to target network node 110-2, UE 120 can continue to send and / or receive data. For example, as part of handover process 700, UE 120 may be configured to switch from communicating using a beam (or beam pair) associated with source network node 110-1 to communicating using a beam (or beam pair) associated with target network node 110-2.

[0114] As elsewhere in this document (such as in conjunction with Figures 4 to 6 ) As described in more detail, source network node 110-1 and / or other devices may utilize machine learning predictions and / or machine learning services to improve communication performance of UE 120 (e.g., make improved beam selection and / or beam blocking predictions). For example, UE 120 may be associated with a vehicle UE that is collecting sensor data and performing pre-processing of the sensor data to obtain machine learning data associated with objects detected near the vehicle UE. The vehicle UE may provide machine learning data to source network node 110-1 and / or an inference host associated with source network node 110-1 (e.g., a source inference host). Source network node 110-1 and / or the source inference host may use the machine learning data (e.g., provided by the vehicle UE) to make predictions for UE 120 using a machine learning model, such as beam selection predictions, beam blocking predictions, and / or other predictions.

[0115] However, when UE 120 is handed over to target network node 110-2 (e.g., as part of handover process 700), target network node 110-2 and / or an inference host associated with target network node 110-2 may not have the benefit of the machine learning data and / or predictions. For example, in some cases, target network node 110-2 may not be able to perform such machine learning operations (e.g., target network node 110-2 may not be associated with an inference host, or an inference host associated with target network node 110-2 may not be able to perform machine learning operations). As a result, communication performance of UE 120 may be degraded because target network node 110-2 may not be able to perform machine learning operations to predict beam blocking and / or select an optimal beam or beam pair for UE 120 to use.

[0116] As another example, even if target network node 110-2 and / or an inference host associated with target network node 110-2 (e.g., a target inference host) is capable of performing machine learning operations, target network node 110-2 and / or the target inference host may have to establish a communication session with UE 120 and with the vehicle UE (e.g., for the exchange of machine learning data from the vehicle UE with the target inference host). Additionally, the machine learning model executed on the target inference host may take some time to converge and provide meaningful and / or accurate predictions. Therefore, there may be a period of time after UE 120 is handed over to target network node 110-2 during which machine learning predictions and / or inaccurate machine learning predictions are not used to select communication parameters (e.g., beams or beam pairs, etc.) for UE 120. Therefore, due to the delay associated with establishing a communication session to exchange machine learning data from the vehicle UE and / or converge the machine learning model used by the target inference host, the communication performance of UE 120 may be degraded.

[0117] Some techniques and apparatus described herein enable UE machine learning service continuity. For example, some techniques and apparatus described herein enable machine learning services associated with a UE to continue while a UE 120 is performing a handover process from a first network node to a second network node. In some aspects, the handover process may include a handover from a source network node to a target network node and from a source inference host to a target inference host.

[0118] For example, in some aspects, machine learning services for UE 120 may have continued through a handover associated with an Xn interface, a midhaul interface, or a backhaul interface. For example, a first network node (e.g., a source network node) may send a handover request communication associated with a handover process for UE 120 to a second network node (e.g., a target network node), wherein (e.g., the source network node) selects the second network node for the handover process based at least in part on first machine learning service information associated with UE 120 or at least one of machine learning capabilities associated with the second network node. As another example, machine learning services for UE 120 may have continued through a handover associated with an N2 interface. For example, a core network node (e.g., an access and mobility function (AMF)) may send a handover request communication associated with a handover process for UE 120 to a second network node (e.g., a target network node), wherein (e.g., the AMF) selects the second network node for the handover process based at least in part on first machine learning service information associated with UE 120 or at least one of machine learning capabilities associated with the second network node.

[0119] A target network node and an inference host associated with the target network node may communicate with a UE (e.g., a vehicle UE) that is providing machine learning data associated with a machine learning service to an end-user UE to establish a communication session (e.g., a protocol data unit (PDU) session) for providing machine learning data associated with the machine learning service to the end-user UE. Additionally, a target inference host associated with the target network node may receive machine learning data and / or machine learning service information associated with a source inference host. The target inference host may make one or more predictions (e.g., an optimal beam or beam pair to be used for communication between the end-user UE and the target network node) based on the machine learning data provided by the UE (e.g., a vehicle UE). The end-user UE may be handed over to the target network node based on the one or more predictions.

[0120] Thus, when an end-user UE is handed over from a first network node to a second network node, machine learning service continuity and machine learning data continuity can be maintained. This enables the target network node to perform a handover process using one or more machine learning predictions associated with the end-user UE. Additionally, the techniques and operations described herein can increase the likelihood that the end-user UE is handed over to a network node that is capable of performing machine learning operations associated with machine learning services. By ensuring machine learning service continuity and machine learning data continuity as part of the handover process, communication performance of a UE (e.g., an end-user UE) can be improved (e.g., because the target network node can quickly utilize accurate machine learning predictions to select communication parameters associated with the UE).

[0121] As indicated above, Figure 7 are provided as examples. Other examples can be found in the Figure 7 Different than described.

[0122] Figure 8 8 is a diagram illustrating an example 800 associated with UE machine learning service continuity according to the present disclosure. Figure 8 As shown, example 800 includes communications between a first UE (e.g., UE 120), a second UE (e.g., UE 120), a source network node (e.g., network node 110 or network node 506, etc.), a source inference host (e.g., near-RT RIC 325, non-RT RIC 315, model inference host 404, or ML inference host 514, etc.), a target network node (e.g., network node 110 or network node 506, etc.), and a target inference host (e.g., near-RT RIC 325, non-RT RIC 315, model inference host 404, or ML inference host 514, etc.). In some aspects, Figure 8The devices and / or nodes depicted in the figure may be included in a wireless network (such as wireless network 100). In some aspects, the first UE and / or the second UE may be in Figure 8 The operations depicted in FIG. 1 have previously established a communication connection with the source network node.

[0123] As used herein, a network node "sending" a communication to a UE may refer to sending directly (e.g., from the network node to the UE) or indirectly via one or more other network nodes or devices. For example, if the network node is a DU, then indirectly sending to the UE may include the DU sending the communication to the RU and the RU sending the communication to the UE 120. Similarly, a UE "sending" a communication to a network node may refer to sending directly (e.g., from the UE to the network node) or indirectly via one or more other network nodes or devices. For example, if the network node is a DU, then indirectly sending to the network node may include the UE sending the communication to the RU and the RU sending the communication to the DU.

[0124] As used herein, a UE "sending" a signal or communicating to an inference host or "receiving" a signal or communicating from an inference host may refer to the UE sending or receiving a signal or communicating via a network node associated with the inference host. For example, the UE may communicate with the inference host via a network node associated with the inference host.

[0125] The source inference host may be associated with a source network node (e.g., machine learning services associated with the source network node may be provided). Similarly, the target inference host may be associated with a target network node (e.g., machine learning services associated with the target network node may be provided). In some aspects, the source inference host and the source network node may be co-located, and / or the target inference host and the target network node may be co-located. In some aspects, the source inference host and the source network node may not be co-located, and / or the target inference host and the target network node may not be co-located. In some aspects, the source inference host may be co-located with or near the CU of the source network node (e.g., geographically). Additionally or alternatively, the target inference host may be co-located with or near the CU of the target network node (e.g., geographically). In some aspects, the source inference host may be co-located with or near the DU or RU of the source network node. Additionally or alternatively, the target inference host may be co-located with or near the DU or RU of the target network node.

[0126] The first UE may be an end user UE. The second UE may be a vehicle UE (eg, may be associated with a vehicle, or included in a vehicle). That is, the second UE may be associated with one or more sensor devices (eg, in combination with a Figure 5 and Figure 6120 can be associated with a vehicle (e.g., a device or machine capable of obtaining sensor data in a manner similar to that described herein). Although the second UE is described herein as being associated with a vehicle, the second UE may be associated with other devices or machines (e.g., a device or machine capable of obtaining sensor data in a manner similar to that described herein). In some aspects, UE 120 may be capable of obtaining sensor data in a manner similar to that described herein. In such examples, the first UE and the second UE may be a single device (e.g., a single UE 120 may perform operations described herein as being performed separately by the first UE and the second UE). In some aspects, the first UE and the second UE may communicate with each other, such as via sidelink communication, Bluetooth communication, Wi-Fi communication, and / or PC5 communication, etc.

[0127] Although some examples described herein are associated with an object detection machine learning data service associated with a second UE (e.g., a vehicle UE), the operations and techniques described herein may be similarly applied to any machine learning service. For example, the machine learning service may be any machine learning service associated with the first UE and / or the second UE that continues (e.g., is handed over) when the first UE and / or the second UE is handed over from a source network node to a target network node.

[0128] In some aspects, as shown by reference numeral 802, the source network node and the target network node may communicate to establish a communication connection via an interface. The interface may be a communication interface between RAN nodes in a wireless network, such as an Xn interface, an X2 interface, a backhaul interface, or a midhaul interface, etc. As part of establishing a communication connection, the source network node and the target network node may exchange ML information. The ML information may indicate one or more ML capabilities. For example, the source network node may send ML information indicating whether the source network node is capable of performing one or more machine learning operations, and the target network node may receive the ML information. Similarly, the target network node may send ML information indicating whether the target network node is capable of performing one or more machine learning operations, and the source network node may receive the ML information. In some aspects, the ML information exchanged between the source network and the target network may include one or more ML services, such as an ML training service and / or an ML inference service. In some aspects, the ML information exchanged between the source network and the target network may include one or more ML training hosts and / or ML inference hosts, and one or more ML models associated with the ML training host or the ML inference host. In some aspects, the ML information exchanged between the source network and the target network may include ML data information for the ML training service and / or the ML inference service, where the ML data information may include one or more ML data types or formats associated with the ML model.

[0129] In some aspects, the ML information may indicate whether the network node is associated with an inference host. For example, a source network node may send an indication that the source network node is associated with at least one inference host (e.g., a source inference host), and the target network node may receive the indication. Similarly, a target network node may send an indication that the target network node is associated with at least one inference host (e.g., a target inference host), and the source network node may receive the indication. In some aspects, the ML information may indicate additional information associated with the ML capabilities of the inference host. For example, the ML information may indicate one or more machine learning models (e.g., a beam prediction model, a handover prediction model, a scheduling model, and / or other machine learning models) supported by the inference host. In this way, the source network node can identify which candidate network nodes (e.g., a target network node) are capable of performing operations to support machine learning services associated with the first UE and / or the second UE.

[0130] For example, the source network node and / or the source inference host may support machine learning services for the first UE and / or the second UE (e.g., in conjunction with the above Figures 4 to 6 For example, the second UE may obtain sensor data (such as one or more sensor measurements or sensor information) via one or more sensor devices. The second UE may obtain sensor data (such as one or more sensor measurements or sensor information) via a machine learning model (e.g., an object detection model, in combination with the above Figure 5 The second UE may (pre)process the sensor data in a manner similar to that described above. The second UE may obtain an output of a machine learning model associated with object detection based at least in part on the sensor data (e.g., which may be referred to herein as machine learning data). For example, the second UE may process the sensor data to identify one or more objects near the second UE and / or identify information associated with the one or more objects (e.g., geographic location, speed, and / or direction of travel, etc.). The second UE may send (e.g., to a source network node and / or to a source inference host) machine learning data associated with object detection predictions for machine learning services. Additionally, the second UE may send (e.g., to a source network node and / or to a source inference host) UE information associated with the first UE or the second UE for machine learning services (e.g., location, speed, direction of travel, orientation, and / or beam information, etc.). The source inference host may use the machine learning data and UE information as input to a machine learning model executed on the source inference host. The machine learning model may output one or more predictions associated with the first UE and / or the second UE, such as a beam prediction or a beam blocking prediction for the first UE (e.g., in conjunction with the above). Figure 5 and Figure 6In this manner, the source network node and / or the source inference host may use the machine learning data and UE information provided by the second UE to support a machine learning service for an end-user UE (e.g., the first UE). For example, the machine learning model associated with the source inference host may be in Figure 8 The operations depicted in are previously trained and / or operable for machine learning services associated with the first UE and / or the second UE.

[0131] As used herein, "machine learning data service" may refer to a machine learning operation performed by a second UE (e.g., to generate machine learning data). "Machine learning inference service" may refer to a machine learning operation performed by an inference host (such as a source inference host or a target inference host).

[0132] As shown by reference numeral 804, the first UE may send a measurement report. Additionally or alternatively, as shown by reference numeral 806, the second UE may send a measurement report. The source network node may receive a measurement report associated with the first UE and / or the second UE. The measurement report may indicate one or more measurements of a candidate cell or a candidate network node (e.g., including a target network node). For example, the first UE and / or the second UE may be configured (e.g., by the source network node or another network node) to perform (e.g., periodically or triggered by an event or condition (such as machine learning performance monitoring)) measurements of a serving network node (e.g., a source network node) and / or one or more candidate network nodes (e.g., including a target network node). For example, these measurements may include RSRP measurements, RSSI measurements, signal strength measurements, signal quality measurements, and / or measurements based on machine learning performance monitoring, and the like.

[0133] As indicated by reference numeral 808, the source network node may determine that a handover (HO) procedure should be initiated for the first UE and / or the second UE. For example, the source network node may determine to initiate the handover procedure based at least in part on a measurement value associated with the target network node and / or the source network node. For example, the source network node may determine that the measurement value associated with the source network node is less than or equal to a threshold. Additionally or alternatively, the source network node may determine that the measurement value associated with the target network node and / or one or more other candidate network nodes is greater than the measurement value associated with the source network node (e.g., a threshold amount), wherein the target network node may be associated with at least one inference host that may support ML services provided by the source inference host to the first UE or the second UE and ML data services provided by the second UE. Thus, the source network node may initiate a handover procedure for the first UE and / or the second UE.

[0134] As shown by reference numeral 810, the source network node may send a handover communication, and the source inference host may receive the handover communication. In some aspects, the handover communication may be a handover indication communication. The handover communication may indicate information associated with a machine learning service for the first UE and / or the second UE. For example, the information associated with the machine learning service may include UE information or UE context, such as an identifier associated with the first UE and / or the second UE (e.g., a UE identifier), a geographic location associated with the first UE and / or the second UE, a direction of travel associated with the first UE and / or the second UE, a speed associated with the first UE and / or the second UE, beam information associated with the first UE and / or the second UE, an orientation of the first UE relative to the second UE, and / or a direction of travel, and the like. Additionally or alternatively, the information associated with the machine learning service may include the machine learning capabilities of the second UE (e.g., one or more machine learning features supported by the second UE) and / or machine learning data service information provided by the second UE, and the like. For example, the information associated with the machine learning service may include an indication of a machine learning model (e.g., a machine learning model type, name, identifier, and / or index associated with one or more functions and / or function structures having a parameter set, which may be preconfigured or configurable or updateable based on offline or online training or inference monitoring) associated with a machine learning service provided to the first UE and / or the second UE (e.g., a first machine learning service type, name, identifier, and / or index associated with a parameter set including one or more QoS profiles or QoS flows) executed on the second UE (e.g., for feature extraction), and / or a machine learning data service (e.g., a second machine learning service type, name, identifier, and / or index having a parameter set including one or more QoS profiles or QoS flows, ML data type or format) that provides machine learning data (e.g., ML inference data) via feature extraction (e.g., object detection) performed by the second UE for the machine learning service for the first UE and / or the second UE, etc. The information associated with the machine learning service provided by the second UE may be referred to herein as "machine learning service information".

[0135] As shown by reference numeral 812, the source inference host may send a handover communication indicating machine learning inference information associated with a machine learning service for the first UE and / or the second UE, and the source network node may receive the handover communication. As described elsewhere herein, "machine learning inference information" may refer to information associated with or provided by an inference host. "Machine learning service information" may refer to information associated with or provided by a UE. In some aspects, the handover communication may be a handover response communication. The handover response communication may be responsive to a handover indication communication sent by the source network node (e.g., as described in conjunction with reference numeral 810). The machine learning inference information may include an indication of a machine learning model (e.g., a machine learning model type, name, identifier, and / or index associated with one or more functions and / or functional structures having a set of parameters, which may be preconfigured or configurable or updateable based on offline or online training or inference monitoring) associated with a machine learning service (e.g., a first machine learning service type, name, identifier, and / or index as described above in conjunction with reference number 810) executed on a source inference host, a predicted state associated with a machine learning service for the first UE and / or the second UE, machine learning inference data information provided by the second UE (e.g., associated with a second machine learning service type, name, identifier, and / or index as described above in conjunction with reference number 810), and the like.

[0136] For example, the handover indication communication may indicate to the source inference host that the first UE and / or the second UE will be handed over to a different network node. In addition to the machine learning information associated with the first UE and / or the second UE, the handover indication communication may also identify the first UE and / or the second UE so that the source inference host can identify the machine learning inference information associated with the machine learning service for the first UE and / or the second UE and / or the machine learning information associated with the machine learning data service provided by the second UE. The handover response communication may provide information related to the machine learning operations performed by the source inference host in conjunction with the machine learning service and / or machine learning data provided by the second UE for the machine learning operations performed by the source inference host. This may enable the source network node to make an improved handover determination for the first UE and / or the second UE (e.g., considering information associated with the machine learning service and the machine learning data service to increase the likelihood of continuity associated with the machine learning service).

[0137] For example, as shown by reference numeral 814, the source network node may select a target network node and / or a target inference host for the handover process. The source network node may select the target network node and / or the target inference host based at least in part on the machine learning capabilities of the second UE, the machine learning data service information (e.g., as described above in conjunction with reference numeral 810), the machine learning inference information (e.g., received from the source inference host as described above in conjunction with reference numeral 812), and / or the machine learning capabilities associated with the target network node (e.g., received by the source network node as described above in conjunction with reference numeral 802), etc. In other words, when selecting a target network node associated with the target inference host for the handover process, the source network node may consider the information associated with the machine learning service. For example, the source network node may select the target network node (and the associated target inference host) based at least in part on the received machine learning information (e.g., the machine learning data service information of the second UE and the machine learning inference information of the source inference host).

[0138] For example, the source network node may identify a candidate network node associated with the best measurement value (e.g., indicated by the first UE and / or the second UE). The source network node may determine whether the identified candidate network node is capable of performing machine learning operations and / or whether the identified candidate network node is capable of performing machine learning services associated with the first UE and / or the second UE. If the identified candidate network node is capable of performing machine learning operations and / or is capable of performing machine learning services, the source network node may select the candidate network node as a target network node for the handover process. However, if the identified candidate network node is not capable of performing machine learning operations and / or is not capable of performing machine learning services, the source network node may not select the candidate network node as a target network node for the handover process. For example, the candidate network node may be associated with a better or greater measurement value than the target network node (e.g., as indicated by a measurement report received by the source network node as described above in conjunction with reference numerals 804 and / or 806). However, the candidate network node may not be able to support machine learning services associated with the first UE and / or the second UE. Therefore, the source network node may avoid selecting the candidate network node as a target network node for the handover process. This may increase the likelihood that machine learning services and / or machine learning data services may be continued for the first UE and / or the second UE after the handover procedure is performed.

[0139] As shown by reference numeral 816, the source network node may send a handover request communication associated with a handover process for the first UE and / or the second UE, and the target network node may receive the handover request communication. In some aspects, the handover request communication may be sent via an Xn interface. The handover request communication may indicate machine learning information received by the source network node (e.g., machine learning data service information of the second UE and machine learning inference information of the source inference host). For example, the handover request communication may indicate ML inference information (e.g., received from the source inference host as described above in conjunction with reference numeral 812) and / or ML data service information associated with the second UE, etc. This may enable the target network node to identify information associated with machine learning services to be supported by the target inference host.

[0140] As indicated by reference numeral 818, the target network node may send a handover communication indicating some or all of the machine learning information indicated in the handover request communication received from the source network node, and the target inference host may receive the handover communication. In some aspects, the handover communication may be a handover indication communication. In some aspects, the handover communication may indicate machine learning information associated with the first UE and / or the second UE. The handover communication to the target inference host may indicate machine learning data service information of the second UE and / or machine learning inference information associated with the source inference host, and so on.

[0141] As shown by reference numeral 820, the target inference host may send a handover communication indicating whether the target inference host is capable of supporting the machine learning service, and the target network node may receive the handover communication. In some aspects, the handover communication may be a handover response communication. For example, the target inference host may determine whether the target inference host is capable of supporting the machine learning service based at least in part on the machine learning information indicated in the handover indication communication received from the target network node. For example, the target inference host may determine whether the target inference host supports a machine learning model associated with the machine learning service. For example, if the machine learning service is associated with beam prediction, the target inference host may determine whether the target inference host supports the beam prediction machine learning model. For example, the target inference host may determine whether the target inference host supports a machine learning data service provided by the second UE. For example, if the machine learning data service is associated with an extracted feature (e.g., an object detected in a certain data type or format), the target inference host may determine whether the target inference host supports the machine learning data service for its machine learning inference model.

[0142] If the target inference host cannot support the machine learning service and / or the machine learning data service, the handover response communication may indicate that the handover process cannot be completed. Alternatively, if the target inference host supports the machine learning service using the machine learning data service provided by the second UE, the handover response communication may indicate that the handover process can continue. In this way, the target network node can communicate with the target inference host to determine that the target inference host is able to support the machine learning service using the machine learning data service provided by the second UE.

[0143] If the target inference host determines that the machine learning service is supported, the target inference host may include machine learning inference information associated with the target inference host. For example, the handover response communication sent to the target network node may include: a target inference host identifier, a machine learning model or a machine learning inference model associated with a machine learning service or a machine learning inference service (e.g., a first machine learning service type, name, identifier, and / or index as described above in conjunction with reference number 810) executed on the target inference host (e.g., a machine learning model type, name, identifier, and / or index associated with one or more functions and / or functional structures having a set of parameters, which may be preconfigured or may be configured or updated based on offline or online training or inference monitoring), machine learning inference information or machine learning data (e.g., a machine learning data service type, name, identifier, and / or index having a set of parameters including one or more QoS profiles or QoS flows, ML data types or formats), and / or other machine learning data associated with the machine learning service associated with the target inference host.

[0144] As shown by reference numeral 822, the target network node may send a handover confirmation communication, and the source network node may receive the handover confirmation communication. In some aspects, the handover confirmation communication may be sent via an Xn interface. The handover confirmation communication may indicate whether the handover is approved or accepted by the target network node. For example, if the target inference host indicates that machine learning services are not supported (e.g., in a handover response communication received by the target network node, as described above in conjunction with reference numeral 820), the handover confirmation communication sent to the source network node may indicate that the handover process is rejected. Alternatively, if the target inference host indicates that machine learning services are supported (e.g., in a handover response communication received by the target network node, as described above in conjunction with reference numeral 820), the handover confirmation communication sent to the source network node may indicate that the handover process is approved.

[0145] Additionally, the handover confirmation communication may indicate machine learning inference information associated with the target inference host. For example, the handover confirmation communication sent to the source network node may include an indication of a machine learning model associated with the machine learning service executed on the target inference host, a machine learning inference host identifier associated with the target inference host, and / or machine learning inference data information, etc.

[0146] As indicated by reference numeral 824, the source network node may send a handover confirmation communication based at least in part on the source network node receiving the handover confirmation communication from the target network node, and the source inference host may receive the handover confirmation communication. For example, the source network node may indicate to the source inference host that the handover of the first UE and / or the second UE is confirmed and / or in progress.

[0147] As shown by reference numeral 826, the source inference host may send machine learning data (e.g., machine learning inference data provided by the second UE) and / or machine learning inferences (e.g., predictions provided by the source inference host) for a machine learning inference service or machine learning service associated with the first UE and / or the second UE, and the source network node may receive the machine learning data and / or the machine learning inference. For example, based at least in part on receiving the handover confirmation communication, the source inference host may provide the machine learning data or machine learning inferences for the machine learning inference service or machine learning service associated with the first UE and / or the second UE to the source network node. As shown by reference numeral 828, the source network node may send machine learning data and / or machine learning inferences for the machine learning inference service or machine learning service associated with the first UE and / or the second UE, and the target network node may receive the machine learning data and / or the machine learning inference. For example, the source network node may forward (e.g., to the target network node) machine learning data buffered at the source inference host or machine learning data from the second UE and / or machine learning inferences generated at the source inference host (e.g., using machine learning data from the second UE).

[0148] In some aspects, an ML data session (e.g., associated with an ML model executed on a second UE) or an ML inference data session (e.g., associated with an ML model executed on an inference host) may be initiated by a network node (e.g., a target network node), as described in conjunction with reference numerals 830 to 842. Additionally or alternatively, the ML data session or ML inference data session may be initiated by a UE (e.g., a second UE) when the UE establishes a connection with a target network node, as described in conjunction with reference numerals 844 to 856.

[0149] As indicated by reference numeral 830, the source network node may send information associated with a machine learning context of a first UE and / or a second UE and machine learning data session information associated with the second UE, and the target network node may receive the information and the machine learning data session information. For example, the UE ML context may indicate UE information (e.g., a first UE identifier, a second UE identifier, and / or UE ML capabilities, etc.) associated with a machine learning inference service or a machine learning service to be maintained by a target network node associated with the first UE and / or the second UE and / or associated with a machine learning inference data service provided by the second UE. For example, the machine learning data session information associated with the second UE may indicate a configuration associated with the machine learning data session. The machine learning data session may include one or more radio bearers (e.g., a signaling radio bearer (SRB) and a data radio bearer (DRB)) and / or one or more PDU sessions with associated parameters (such as a QoS profile or a QoS flow or a 5G QoS identifier (5QI)). In some aspects, if the inference host is allocated in a logical architecture at a layer higher than the communication protocol stack of the network node (e.g., at a sublayer or adaptation layer for machine learning services), one or more radio bearers (e.g., DRBs with associated QoS flows) and / or PDCP entities and / or logical channels may be established by the network node for the network node user plane.

[0150] In such examples, the machine learning data session information may include information associated with the radio bearer (e.g., a radio bearer identifier, a 5QI, a QoS flow, or a QoS profile with QoS parameters and QoS characteristics, etc.). In some aspects, if the inference host is anchored at a local user plane function (UPF) (e.g., which is co-located with the network node), one or more PDU sessions may be initiated by the network node. In this case, the machine learning data session information may include information associated with the PDU session (e.g., a PDU session identifier, a PDU session type, service and session continuity modes, user plane security information, a slice identifier for machine learning, and / or a QoS flow based on a 5QI or a QoS profile, etc.).

[0151] In some aspects, the target network node and the target inference host may communicate to establish a machine learning data session to be used to exchange machine learning data with the second UE (e.g., establish a user plane for data traffic), such as when the target inference host is co-located with the target network node. For example, the target network node may handle PDU session establishment (e.g., when the PDU session is anchored at a local UPF connected to the target inference host). In some other aspects, the core network node may handle PDU session establishment (e.g., the core network node may switch a PDU session associated with the second UE from being associated with a source inference host to being associated with a target inference host), such as when the target inference host is not co-located with the target network node. In such examples, the target network node or the source network node may send a request to switch a PDU session associated with the second UE from the source inference host to the target inference host, and the core network node may receive the request.

[0152] As indicated by reference numeral 832, the target network node may send a first data session communication, and the target inference host may receive the first data session communication, the first data session communication requesting one or more data sessions for a machine learning data service for a machine learning service for the first UE and / or the second UE provided by the second UE. For example, the first data session communication may be used to establish (with an associated QoS flow mapping) a radio bearer and / or PDCP entity with a machine learning sublayer or adaptation layer, where the target inference host is assigned for the machine learning service. For example, the first data session communication may be a PDU establishment or PDU request communication with a local UPF, where the target inference host is anchored. The first data session communication may indicate (e.g., based at least in part on the UE ML context and ML data session information as described above in conjunction with reference numeral 830) a second UE (e.g., an identifier of the second UE), a machine learning data service (e.g., a machine learning data service type, name, identifier, and / or index), one or more ML data sessions to be established using the associated ML data session information, and the like.

[0153] As indicated by reference numeral 834, the target inference host may send, and the target network node may receive, a second data session communication indicating one or more data sessions for a machine data learning service provided by the second UE associated with a machine learning service for the first UE and / or the second UE. For example, the second data session communication may be a response to a radio bearer establishment, a PDU establishment, or a PDU request communication (e.g., as described above in conjunction with reference numeral 832), wherein the response may confirm the data session establishment or deny the data session establishment. To confirm the data session establishment, the second data session communication may indicate, among other things, one or more ML data session identifiers (e.g., radio bearers or PDU identifiers) associated with the machine learning data service provided by the second UE and an associated ML data session type (e.g., PDU or non-PDU, IP or non-IP, structured or unstructured, etc.).

[0154] As indicated by reference numeral 836, the target network node may send a third machine learning data session communication associated with the machine learning data service, and the source network node may receive the third machine learning data session communication. For example, the third machine learning data session communication may indicate a machine learning data session modification of a machine learning data service provided by the second UE for the machine learning service for the first UE and / or the second UE based at least in part on the second data session communication (e.g., a response to a radio bearer establishment or a PDU establishment or a PDU request communication for confirming the data session establishment, as described above in conjunction with reference numeral 834). The machine learning data session modification may include information related to the machine learning data session established as described above in conjunction with reference numeral 834. For example, the source network node may receive one or more ML data session identifiers and associated ML data session types for the machine learning data service provided by the second UE, an identifier of the target inference host, an indication of resources associated with the data session (e.g., radio resources or other network resources), and / or data session configuration parameters, etc. In some aspects, the source network node may trigger a release of the ML data session between the source inference host and the second UE for the machine learning data service provided to the source inference host.

[0155] As shown by reference numeral 838, the target network node may send machine learning data and / or machine learning inferences associated with the machine learning service, and the target inference host may receive the machine learning data and / or the machine learning inferences. For example, the machine learning data and / or machine learning inferences associated with the machine learning service for the first UE and / or the second UE (e.g., the information described above in conjunction with reference numeral 828) forwarded by the source network node may be sent via a user plane established using an ML data session establishment (e.g., as described above in conjunction with reference numerals 832 and 834). This may enable the target inference host to collect machine learning data and / or machine learning inferences to be used to train and / or aggregate machine learning models associated with the machine learning service for the first UE and / or the second UE. For example, the target inference host may use machine learning data (e.g., machine learning data collected by the second UE) and / or machine learning inferences generated by the source inference host to train and / or aggregate machine learning models to be used to support machine learning services for the first UE and / or the second UE. This may reduce the amount of time and / or processing overhead associated with training machine learning models, among other things.

[0156] In some aspects, a second UE (e.g., a UE that collects and / or provides machine learning data) may be handed over to a target network node before a first UE (e.g., an end-user UE). This may enable the second UE to provide machine learning data to a target inference host so that the target inference host can make one or more predictions (e.g., beam or beam pair predictions) using the machine learning data for handing over the first UE. For example, as shown by reference numeral 840, the source network node may send a handover command communication indicating that the second UE is to perform a handover with the target network node, and the second UE may receive the handover command communication. The handover command communication may be an RRC communication. The handover command may indicate a target network node (e.g., an identifier associated with the target network node), a target inference host (e.g., an identifier associated with the target inference host), machine learning data session information (e.g., one or more machine learning data session identifiers and related machine learning data session parameters, such as a machine learning data session type), machine learning data service information (e.g., a machine learning service type, name, identifier, and / or index as described above in conjunction with reference number 810), and / or other communication parameters to be used by the second UE to establish a connection with the target network node and / or the target inference host.

[0157] The handover command communication may indicate machine learning inference information associated with a target inference host associated with the target network node. For example, the handover command communication may indicate a machine learning model associated with a machine learning service executed on the target inference host, a machine learning inference host identifier associated with the target inference host, and / or machine learning data session information, etc. As part of the handover process, this may enable the second UE to identify the machine learning information associated with the target network node. For example, the second UE may identify an identifier of the target inference host, a machine learning model associated with the target inference host, and / or machine learning data session information associated with the target inference host.

[0158] As indicated by reference numeral 842, the second UE may send a communication for establishing a communication connection (e.g., an RRC connection) with the target network node, and the target network node may receive the communication. For example, the second UE may send a RACH communication (e.g., msgA or msg1) or an RRC connection communication. The communication may indicate machine learning data session information associated with a machine learning data service provided by the second UE (e.g., a data session identifier and associated QoS parameters), information associated with the machine learning data (e.g., a type, format, or structure), information associated with the machine learning data service (e.g., information associated with a machine learning model associated with the machine learning service executed on the second UE, an indication of one or more machine learning service types associated with a QoS profile or QoS flow or 5QI, one or more machine learning model identifiers associated with the machine learning service, and / or one or more machine learning inference host identifiers associated with the machine learning service), and the like. For example, the communication may indicate the machine learning data service information as part of establishing an RRC connection with the target network node.

[0159] In some aspects, an ML data session or an ML inference data session may be initiated by a UE (e.g., a second UE) when the UE establishes a connection with a target network node. For example, as indicated by reference numeral 844, the source network node may send information associated with a machine learning context of a first UE and / or a second UE and machine learning data service information associated with the second UE, and the target network node may receive the information and the machine learning data service information. For example, the UE ML context may indicate UE information (e.g., a first UE identifier and / or a second UE identifier and / or a UE ML capability, etc.) associated with a machine learning inference service or a machine learning service to be maintained by a target network node associated with the first UE and / or the second UE and / or associated with a machine learning inference data service provided by the second UE.

[0160] As indicated by reference numeral 846, the source network node may send a handover command communication indicating that the second UE is to perform a handover with the target network node, and the second UE may receive the handover command communication. The handover command communication may be an RRC communication. The handover command may indicate a target network node (e.g., an identifier associated with the target network node), a target inference host (e.g., an identifier associated with the target inference host), machine learning data service information (e.g., a machine learning data service type, name, identifier, and / or index as described above in conjunction with reference numeral 810), and / or other communication parameters to be used by the second UE to establish a connection with the target network node to continue to provide the target inference host with machine learning data services for machine learning services associated with the first UE and / or the second UE.

[0161] The handover command communication may indicate machine learning inference information associated with a target inference host associated with the target network node. For example, the handover command communication may indicate a machine learning model associated with a machine learning service executed on the target inference host, and / or a machine learning inference host identifier associated with the target inference host, machine learning data (e.g., data type, format, or structure), etc. As part of the handover process, this may enable the second UE to indicate machine learning information associated with the target network node. For example, the second UE may indicate an identifier of the target inference host, a machine learning model associated with the target inference host, and / or machine learning data for initiating an ML data session establishment with the target inference host.

[0162] As shown by reference numeral 848, the second UE may send a communication for establishing a communication connection (e.g., an RRC connection) with the target network node, and the target network node may receive the communication. For example, the second UE may send a RACH communication (e.g., msgA or msg1) or an RRC connection communication. The communication may include an identification associated with a machine learning data service (e.g., indicating a machine learning data service type, name, identifier, and / or index), UE machine learning capabilities, information associated with machine learning data (e.g., type, format, or structure), information associated with the machine learning data service (e.g., information associated with a machine learning model associated with the machine learning service executed on the second UE, an indication of one or more machine learning service types associated with a QoS profile or a QoS flow, one or more machine learning model identifiers associated with the machine learning service, and / or one or more machine learning inference host identifiers associated with the machine learning service), and the like. For example, the machine learning data service information may be used for the second UE to initiate the establishment of a machine learning data session with the target inference host (e.g., using the machine learning data service information and / or the machine learning inference information as described above in conjunction with reference numeral 846 to indicate an ML data session establishment request to the target network node).

[0163] As shown by reference numeral 850, based on the ML data session establishment request received from the second UE, the target network node may send a first data session communication, and the target inference host may receive the first data session communication, the first data session communication indicating one or more data sessions for machine learning data services provided by the second UE for machine learning services for the first UE and / or the second UE. For example, the first data session communication may be used to establish (with associated QoS flow mapping) a radio bearer and / or PDCP entity with a machine learning sublayer or adaptation layer, where the target inference host is assigned for the machine learning service. For example, the first data session communication may be a PDU establishment or PDU request communication with a local UPF at which the target inference host is anchored. The first data session communication may indicate (e.g., based at least in part on the UE ML context as described above in conjunction with reference numeral 844 and the ML data service information and / or machine learning inference information as described above in conjunction with reference numeral 848) a second UE (e.g., an identifier of the second UE), a machine learning data service (e.g., a machine learning data service type, name, identifier, and / or index), one or more ML data sessions to be established using the ML data service information and / or machine learning inference information, and the like.

[0164] As indicated by reference numeral 852, the target inference host may send, and the target network node may receive, a second data session communication confirming one or more data sessions for a machine data learning service provided by the second UE for a machine learning service for the first UE and / or the second UE. For example, the second data session communication may be a response to a radio bearer establishment or a PDU establishment or a PDU request communication (e.g., as described above in conjunction with reference numeral 850), wherein the response may confirm the data session establishment or deny the data session establishment. To confirm the data session establishment, the second data session communication may indicate, among other things, one or more ML data session identifiers (e.g., radio bearers or PDU identifiers) associated with the machine learning data service provided by the second UE and an associated ML data session type (e.g., PDU or non-PDU, IP or non-IP, structured or unstructured, etc.).

[0165] As indicated by reference numeral 854, the target network node may send a third machine learning data session communication associated with the machine learning data service, and the source network node may receive the third machine learning data session communication. For example, the machine learning data session communication may indicate a machine learning data session modification of a machine learning data service for a first UE and / or a second UE provided by the second UE based at least in part on a second data session communication (e.g., a response to a radio bearer establishment or a PDU establishment or a PDU request communication used to confirm the data session establishment, as described above in conjunction with reference numeral 852). In some aspects, the source network node may trigger a release of the ML data session between the source inference host and the second UE for use in the machine learning data service provided to the source inference host.

[0166] As shown by reference numeral 856, the target network node may send machine learning data and / or machine learning inferences associated with the machine learning service, and the target inference host may receive the machine learning data and / or the machine learning inferences. For example, the machine learning data and / or machine learning inferences associated with the machine learning service for the first UE and / or the second UE (e.g., forwarded by the source network node as described above in conjunction with reference numeral 828) may be sent via a user plane established using an ML data session establishment (e.g., as described above in conjunction with reference numerals 850 and 852). This may enable the target inference host to collect machine learning data and / or machine learning inferences to be used to train and / or aggregate machine learning models associated with the machine learning service for the first UE and / or the second UE. For example, the target inference host may use machine learning data (e.g., machine learning data collected by the second UE) and / or machine learning inferences generated by the source inference host to train and / or aggregate machine learning models to be used to support the machine learning service for the first UE and / or the second UE. This may reduce the amount of time and / or processing overhead associated with the source inference host training the machine learning model, among other things.

[0167] In some aspects, a second UE (e.g., a UE that collects and / or provides machine learning data) can be handed over to a target network node before a first UE (e.g., an end-user UE). This can enable the second UE to provide machine learning data or machine learning inference data to a target inference host so that the target inference host can make one or more predictions (e.g., beam or beam pair predictions) using the machine learning data used to hand over the first UE.

[0168] Thus, an ML data session may be established between the second UE and the target inference host (e.g., initiated by the target network node as described above in conjunction with reference numeral 832 or initiated by the second UE as described above in conjunction with reference numeral 848), as shown by reference numeral 858. The second UE may use the ML data session to send machine learning data (e.g., generated by the second UE) to the target inference host. For example, as part of a handover process, a communication connection (e.g., an RRC connection) may be switched from the source network node to the target network node (e.g., as described above such as in conjunction with reference numerals 846 and 848). Additionally, a machine learning data connection (e.g., a user plane connection) may be switched from the source inference host to the target inference host. This may enable the second UE to communicate with the target network node and provide machine learning data to the target inference host.

[0169] For example, as indicated by reference numeral 860, the second UE may send second machine learning data for a machine learning service (e.g., based at least in part on receiving a handover command communication to handover the second UE to the target network node and / or based at least in part on establishing an ML data session), and the target inference host may receive the second machine learning data. For example, the second UE may process one or more sensor measurements via a machine learning model executed on the second UE to obtain second machine learning data associated with the machine learning service. The second UE may send the second machine learning data to the second network node or the second inference host via the ML data session. This may enable the target inference host to collect machine learning data that may be used to perform machine learning operations associated with the second UE and / or the first UE (e.g., before the first UE is handed over to the target network node).

[0170] For example, the target inference host may process machine learning data (e.g., received via an ML data session or received during a handover process, as described in more detail elsewhere herein) to train and / or converge a machine learning model executed on the target inference host. The target inference host may process the machine learning data provided by the second UE to perform one or more machine learning operations associated with the first UE. For example, the target inference host may generate one or more inferences or predictions associated with communication parameters to be used by the first UE to communicate with the target network node. For example, the target inference host may use the machine learning model to determine an optimal beam, an optimal beam pair, an optimal time / frequency resource, an optimal transmit power, and / or an optimal MCS, etc., to be used by the first UE to communicate with the target network node.

[0171] As indicated by reference numeral 862, the target inference host may send one or more machine learning predictions associated with the first UE (e.g., where the one or more machine learning predictions are based at least in part on machine learning data generated by a second UE and provided to the target inference host), and the target network node may receive the one or more machine learning predictions. As indicated by reference numeral 864, the target network node may send an indication of one or more machine learning predictions associated with the first UE, and the source network node may receive the indication. For example, the target network node may send the indication of the one or more machine learning predictions via an Xn interface. The source network node and the target network node may proceed with performing a handover procedure for the first UE based on the one or more predictions (e.g., by configuring the first UE to communicate with the target network node using one or more communication parameters or beams indicated by the one or more predictions).

[0172] For example, the source network node may send UE context information and / or data associated with the first UE, and the target network node may receive UE context information and / or data associated with the first UE, as indicated by reference numeral 866. For example, the UE context information may include an identifier associated with the first UE and / or other information associated with the first UE (e.g., machine learning capabilities or machine learning information) to be stored by the target network node.

[0173] As indicated by reference numeral 868, the source network node may send a handover command communication, and the first UE may receive the handover command communication. For example, the source network node may send an RRC configuration message to the first UE (e.g., in an RRC reconfiguration message) by forwarding an RRC configuration associated with the target network node to the first UE. For example, the target network node may indicate an RRC configuration to be used by the first UE in a handover confirmation communication (e.g., sent as described above in connection with reference numeral 822). For example, the RRC configuration indicated in the handover command communication may be based at least in part on one or more predictions generated by the target inference host (e.g., sent as described above in connection with reference numeral 864).

[0174] The first UE may change the RRC connection from the source network node to the target network node based at least in part on the handover command. As shown in reference numeral 870, the first UE may send a RACH or RRC connection message to the target network node based on the prediction indicated in the handover command from the source network node. The RRC connection message may indicate that the first UE has changed the RRC connection from the source network node to the target network node. In some aspects, the target network node may send a UE context release message to the source network node. The UE context release message may indicate that the handover of the first UE to the target network node is successful.

[0175] The first UE may communicate with the target network node based at least in part on performing the handover procedure. For example, the first UE may communicate with the target network node using one or more communication parameters or beams indicated by one or more predictions generated by the target inference host. Thus, machine learning services associated with the first UE may continue immediately when the first UE is handed over to the target network node (e.g., without delays associated with establishing an ML data session for the second UE to provide machine learning data to the target inference host or training a machine learning model executed on the target inference host).

[0176] Thus, when the first UE is handed over from the source network node to the target network node, machine learning service continuity and machine learning data continuity can be maintained. This enables the target network node to perform a handover process using one or more machine learning predictions associated with the first UE. Additionally, the handover process described herein can increase the likelihood that the first UE is handed over to a network node capable of performing machine learning operations associated with the machine learning service. By ensuring machine learning service continuity and machine learning data continuity as part of the handover process, communication performance of the first UE (e.g., an end user UE) can be improved (e.g., because the target network node can quickly utilize accurate machine learning predictions to select communication parameters associated with the first UE).

[0177] As indicated above, Figure 8 are provided as examples. Other examples can be found in the Figure 8 Different than described.

[0178] Fig. 9A and Fig. 9B is a diagram illustrating an example 900 associated with UE machine learning service continuity according to the present disclosure. Fig. 9A and Fig. 9B As shown, example 900 includes a first UE (eg, UE 120 or Figure 8 ), a second UE (e.g., UE 120 or Figure 8 , a second UE depicted in FIG. 1 ), a source network node (eg, network node 110, network node 506, or Figure 8 ), a source inference host (e.g., a near-RT RIC 325, a non-RT RIC 315, a model inference host 404, an ML inference host 514, or Figure 8 , etc.), a target network node (e.g., network node 110, network node 506, or Figure 8 ), a target inference host (e.g., a near-RT RIC 325, a non-RT RIC 315, a model inference host 404, an ML inference host 514, or Figure 8 ) and one or more core network nodes (e.g., shown as AMF and Session Management Function (SMF)).

[0179] As shown, the source network node and the target network node may be associated with a user plane function (UPF). For example, the source network node may be associated with a local UPF (e.g., a UPF co-located with the source network node), and the target network node may be associated with a local UPF (e.g., a UPF co-located with the target network node). In some aspects, Fig. 9A and Fig. 9B The devices and / or nodes depicted in the figure may be included in a wireless network (such as wireless network 100). In some aspects, the first UE and / or the second UE may be in Fig. 9A The operations depicted in FIG. 1 have previously established a communication connection with the source network node.

[0180] The source network node and / or source inference host may be used in conjunction with other places in this document (such as in conjunction with Figures 4 to 6 and Figure 8 ) may support machine learning services for the first UE and / or the second UE in a manner similar to that described in more detail in Example 800. While example 800 depicts an example of maintaining machine learning continuity during a handover process handled or controlled by a network node, example 900 is associated with an example of maintaining machine learning continuity during a handover process handled or controlled by one or more core network nodes. For example, one or more core network nodes may have improved knowledge about machine learning services or operations supported by various network nodes or inference hosts deployed in a wireless network. For example, the source network node may not receive an indication associated with a machine learning service or operation that the target network node and / or target inference host is capable of supporting. Thus, one or more core network nodes may select a target network node for a handover process associated with the first UE and / or the second UE to increase the likelihood that machine learning service continuity and / or machine learning data continuity can be maintained during and / or after the handover process.

[0181] In some aspects, the core network may include a functional architecture in which the systems and / or methods described herein may be implemented. For example, the core network may include an example architecture of a 5G next generation (NG) core network included in a 5G wireless telecommunications system. Although the example architecture of the core network described herein may be an example of a service-based architecture, in some aspects, the core network may be implemented as a reference point architecture and / or a 4G core network, etc.

[0182] For example, the core network may include multiple functional elements. These functional elements may include, for example, a network slice selection function (NSSF), a network open function (NEF), an authentication server function (AUSF), a unified data management (UDM) component, a policy control function (PCF), an application function (AF), an AMF, an SMF and / or an UPF, etc. These functional elements may be communicatively connected via a message bus. Each of these functional elements may be implemented on one or more devices associated with a wireless telecommunications system. In some specific implementations, one or more of these functional elements may be implemented on a physical device (such as an access point, a network node, a base station and / or a gateway, etc.). In some specific implementations, one or more of these functional elements may be implemented on a computing device in a cloud computing environment.

[0183] The NSSF may include one or more devices that select a network slice instance for a UE. A network slice is a network architecture model in which logically different network slices operate using a common network infrastructure. The NSSF may determine a set of network slice policies to be applied at the network. For example, the NSSF may apply one or more UE routing selection policy (URSP) rules. The NEF may include one or more devices that support the opening of capabilities and / or events (e.g., machine learning capabilities, such as ML data collection, ML training, ML inference, etc.) in a wireless telecommunication system to help other entities in the wireless telecommunication system discover network services. The AUSF may include one or more devices that act as an authentication server and support the process of authenticating a UE in a wireless telecommunication system (e.g., authenticating a machine learning data service provided by a second UE or a machine learning service for a first UE and / or a second UE). The UDM may include one or more devices that store user data and profiles in a wireless telecommunication system. The PCF may include one or more devices that provide a policy framework (e.g., a policy framework for machine learning data services or machine learning services) that combines network slicing, roaming, packet processing, and / or mobility management, etc. The AF may include one or more devices that support application impact on traffic routing, access and / or policy control to the NEF, etc. (e.g., docking with machine learning applications or services).

[0184] The AMF may include one or more devices that act as a termination point for non-access stratum (NAS) signaling and / or mobility management, etc. In some aspects, the AMF may request the NSSF to select a network slice instance for the UE, for example, at least in part in response to a request for a data service from the UE (e.g., a machine learning data service from a second UE). The SMF may include one or more devices that support the establishment, modification, and release of a communication session (e.g., a machine learning data session for a second UE) in a wireless telecommunications system. For example, the SMF may configure service steering policies and / or enforce user equipment IP address allocation and policies at the UPF, etc. In some aspects, the SMF may supply the UE with a network slice instance selected by the NSSF. The UPF may include one or more devices that act as an anchor point for intra-RAT and / or inter-RAT mobility. In some aspects, the UPF may apply rules to packets, such as rules related to packet routing, service reporting, and / or handling of user plane QoS, etc.

[0185] like Fig. 9AAs shown in reference numeral 902, the first UE may send a measurement report. Additionally or alternatively, as shown in reference numeral 904, the second UE may send a measurement report. The source network node may receive a measurement report associated with the first UE and / or the second UE. The measurement report may indicate one or more measurements of a candidate cell or a candidate network node (e.g., including a target network node). For example, the first UE and / or the second UE may be configured (e.g., by the source network node or another network node) to perform (e.g., periodically or triggered by an event or condition (such as machine learning performance monitoring)) measurements of a serving network node (e.g., a source network node) and / or one or more candidate network nodes (e.g., including a target network node). For example, these measurements may include RSRP measurements, RSSI measurements, signal strength measurements, signal quality measurements, and / or measurements based on machine learning performance monitoring, and the like.

[0186] As indicated by reference numeral 906, the source network node may determine that a HO procedure should be initiated for the first UE and / or the second UE. For example, the source network node may determine to initiate a handover procedure based at least in part on a measurement value associated with the target network node and / or the source network node. For example, the source network node may determine that a measurement value associated with the source network node is less than or equal to a threshold. Additionally or alternatively, the source network node may determine that a measurement value associated with the target network node and / or one or more other candidate network nodes is greater than a measurement value associated with the source network node (e.g., a threshold amount), wherein the target network node may be associated with at least one inference host that may support ML services provided by the source inference host to the first UE or the second UE and ML data services provided by the second UE. Thus, the source network node may initiate a handover procedure for the first UE and / or the second UE.

[0187] As shown by reference numeral 908, the source network node may send a handover communication, and the source inference host may receive the handover communication. In some aspects, the handover communication may be a handover indication communication. The handover communication may indicate information associated with a machine learning service for the first UE and / or the second UE. For example, the information associated with the machine learning service may include UE information or UE context, such as an identifier associated with the first UE and / or the second UE (e.g., a UE identifier), a geographic location associated with the first UE and / or the second UE, a direction of travel associated with the first UE and / or the second UE, a speed associated with the first UE and / or the second UE, beam information associated with the first UE and / or the second UE, an orientation or direction of travel of the first UE relative to the second UE. Additionally, the information associated with the machine learning service may include an indication of the machine learning capabilities of the second UE (e.g., one or more machine learning features supported by the second UE) or machine learning data service information provided by the second UE. For example, the information associated with the machine learning service may include an indication of a machine learning model (e.g., a first machine learning service type, name, identifier, and / or index associated with a set of parameters including one or more QoS profiles or QoS flows) associated with the machine learning service provided to the first UE and / or the second UE (e.g., a first machine learning service type, name, identifier, and / or index associated with a set of parameters including one or more QoS profiles or QoS flows), which is executed on the second UE (e.g., for feature extraction), and / or a machine learning data service (e.g., a second machine learning service type, name, identifier, and / or index with a set of parameters including one or more QoS profiles or QoS flows, ML data type or format) associated with providing machine learning data (e.g., ML inference data) via feature extraction (e.g., object detection) performed by the second UE for the machine learning service for the first UE and / or the second UE, and the like. The information associated with the machine learning service provided by the second UE may be referred to herein as "ML inference data service information" or "ML data service information".

[0188] As indicated by reference numeral 910, the source inference host may send a handover communication indicating machine learning inference information associated with the machine learning service for the first UE and / or the second UE, and the source network node may receive the handover communication. In some aspects, the handover communication may be a handover response communication. The handover response communication may be responsive to a handover indication communication sent by the source network node (e.g., as described in conjunction with reference numeral 908). The machine learning inference information may include an indication of a machine learning model (e.g., a machine learning model type, name, identifier, and / or index associated with one or more functions and / or functional structures having a set of parameters, which may be preconfigured or configurable or updateable based on offline or online training or inference monitoring) associated with a machine learning service (e.g., a first machine learning service type, name, identifier, and / or index as described above in conjunction with reference number 908) executed on a source inference host, a predicted state associated with a machine learning service for the first UE and / or the second UE, machine learning inference data information provided by the second UE (e.g., associated with a second machine learning service type, name, identifier, and / or index as described above in conjunction with reference number 908), and the like.

[0189] For example, the handover indication communication may indicate to the source inference host that the first UE and / or the second UE will be handed over to a different network node. In addition to the machine learning information associated with the first UE and / or the second UE, the handover indication communication may also identify the first UE and / or the second UE so that the source inference host can identify the machine learning inference information associated with the machine learning service for the first UE and / or the second UE and / or the machine learning information associated with the machine learning data service provided by the second UE. The handover response communication may provide information related to the machine learning operations performed by the source inference host in conjunction with the machine learning service and / or machine learning data provided by the second UE for the machine learning operations performed by the source inference host. This may enable the source network node to make an improved handover determination for the first UE and / or the second UE (e.g., considering information associated with the machine learning service and the machine learning data service to increase the likelihood of continuity associated with the machine learning service).

[0190] As indicated by reference numeral 912, the source network node may send a handover required communication associated with a handover process for the first UE and / or the second UE, and the core network node (e.g., AMF) may receive the handover required communication. In some aspects, the handover required communication may be sent via the N2 interface. In some aspects, the handover required communication may indicate machine learning information (e.g., inference information associated with a machine learning service for the first UE and / or the second UE, machine learning data associated with a machine learning data service provided by the second UE) and / or an indication of a source inference host associated with the source network node. For example, as part of a request to initiate a handover process for the first UE and / or the second UE, the source network node may send machine learning information associated with machine learning inference information (e.g., as described above in conjunction with reference numeral 910) and machine learning information associated with UE machine learning information (e.g., as described above in conjunction with reference numeral 908) to the AMF. Additionally, the handover required communication may indicate measurement values ​​associated with one or more candidate network nodes (e.g., including a target network node). This may enable the AMF to make an improved handover determination for the first UE and / or the second UE (e.g., taking into account information associated with the machine learning service and the machine learning data service to increase the likelihood of continuity associated with the machine learning service).

[0191] For example, as shown by reference numeral 914, a core network node (e.g., an AMF) may select a target network node and / or a target inference host for the handover process. The core network node may select the target network node and / or the target inference host based at least in part on the machine learning capabilities or machine learning data service information of the second UE (e.g., as described above in conjunction with reference numeral 908) and / or machine learning inference information (e.g., as received from the source inference host as described above in conjunction with reference numeral 910) and / or the machine learning capabilities associated with the target network node (e.g., a candidate target network node managed by the AMF), and the like. In other words, when selecting a target network node for the handover process, the core network node may consider information associated with the machine learning service. For example, the core network node may select the target network node (and the associated target inference host) based at least in part on the received machine learning information (e.g., the machine learning data service information of the second UE and the machine learning inference information of the source inference host).

[0192] For example, the core network node may identify a candidate network node associated with the best measurement value (e.g., as indicated by the first UE and / or the second UE). The core network node may determine whether the identified candidate network node is capable of performing machine learning operations and / or whether the identified candidate network node is capable of performing machine learning services associated with the first UE and / or the second UE. If the identified candidate network node is capable of performing machine learning operations and / or is capable of performing machine learning services, the core network node may select the candidate network node as a target network node for the handover process. However, if the identified candidate network node is not capable of performing machine learning operations and / or is not capable of performing machine learning services, the core network node may not select the candidate network node as a target network node for the handover process (e.g., in a manner similar to that described in more detail elsewhere herein).

[0193] As shown by reference numeral 916, the core network node may send a handover request communication associated with a handover process for the first UE and / or the second UE, and the target network node may receive the handover request communication. In some aspects, the handover request communication may be sent via the N2 interface. The handover request communication may indicate machine learning information received by the core network node (e.g., machine learning data service information of the second UE and machine learning inference service information of the source inference host). For example, the handover request communication may indicate machine learning inference information (e.g., received from the source inference host as described above in conjunction with reference numeral 912) and / or machine learning data service information associated with the second UE, and so on. This may enable the target network node to identify information associated with machine learning services to be supported by the target inference host.

[0194] As indicated by reference numeral 918, the target network node may send a handover communication indicating some or all of the machine learning information indicated in the handover request communication received from the source network node, and the target inference host may receive the handover communication. In some aspects, the handover communication may be a handover indication communication. In some aspects, the handover communication may indicate machine learning information associated with the first UE and / or the second UE. The handover communication to the target inference host may indicate machine learning data service information of the second UE and / or machine learning inference information of the source inference host, and so on.

[0195] As shown by reference numeral 920, the target inference host may send a handover communication indicating whether the target inference host is capable of supporting the machine learning service, and the target network node may receive the handover communication. In some aspects, the handover communication may be a handover response communication. For example, the target inference host may determine whether the target inference host is capable of supporting the machine learning service based at least in part on the machine learning information indicated in the handover indication communication received from the target network node. For example, the target inference host may determine whether the target inference host supports a machine learning model associated with the machine learning service. For example, if the machine learning service is associated with beam prediction, the target inference host may determine whether the target inference host supports the beam prediction machine learning model. For example, the target inference host may determine whether the target inference host supports a machine learning data service provided by the second UE. For example, if the machine learning data service is associated with an extracted feature (e.g., an object detected in a certain data type or format), the target inference host may determine whether the target inference host supports the machine learning data service for its machine learning inference model.

[0196] If the target inference host cannot support the machine learning service and / or the machine learning data service, the handover response communication may indicate that the handover process cannot be completed. Alternatively, if the target inference host supports the machine learning service using the machine learning data service provided by the second UE, the handover response communication may indicate that the handover process can continue. In this way, the target network node can communicate with the target inference host to determine that the target inference host is able to support the machine learning service using the machine learning data service provided by the second UE.

[0197] If the target inference host determines that the machine learning service is supported, the target inference host may include machine learning inference information associated with the target inference host. For example, the handover response communication sent to the target network node may include: a target inference host identifier, a machine learning model or a machine learning inference model associated with a machine learning service or a machine learning inference service (e.g., a first machine learning service type, name, identifier, and / or index as described above in conjunction with reference number 908) executed on the target inference host (e.g., a machine learning model type, name, identifier, and / or index associated with one or more functions and / or functional structures having a set of parameters, which may be preconfigured or may be configured or updated based on offline or online training or inference monitoring), machine learning inference data information or machine learning data (e.g., a machine learning data service type, name, identifier, and / or index having a set of parameters including one or more QoS profiles or QoS flows, ML data types or formats), and / or other machine learning data associated with the machine learning service associated with the target inference host.

[0198] In some aspects, the core network node may communicate to a local UPF associated with the target network node and / or the target inference host for handover preparation. For example, the AMF may send a PDU session update request (e.g., including a PDU session identifier associated with the machine learning data service, an indication of the source network node and / or the source inference host and the associated local UPF, an indication of the target network node identifier and / or the target inference host identifier and the associated local UPF, a UE ML context, etc.) to the SMF based at least in part on a handover need message received from the source network node (e.g., as described above in conjunction with reference number 912) and a target network node selection and / or a target inference host selection (e.g., as described above in conjunction with reference number 914). Based at least in part on the received PDU session update request message, the SMF may transmit a PDU session modification request to a local UPF associated with the source network node and / or the source inference host, and the SMF may further transmit a PDU session establishment request to a local UPF associated with the target network node and / or the target inference host. After receiving a PDU session modification response from a local UPF associated with the source network node and / or the source inference host and receiving a PDU session establishment response from a local UPF associated with the target network node and / or the target inference host, the SMF may respond to the AMF with a PDU session update request response message.

[0199] As shown by reference numeral 922, the target network node may send a handover confirmation (ACK) communication, and the core network node (e.g., AMF) may receive the handover confirmation (ACK) communication. In some aspects, the handover confirmation communication may be sent via the N2 interface. The handover confirmation communication may indicate whether the handover is approved or accepted by the target network node. For example, if the target inference host indicates that machine learning services are not supported (e.g., in a handover response communication received by the target network node, as described above in conjunction with reference numeral 920), the handover confirmation communication sent to the source network node may indicate that the handover process is rejected. Alternatively, if the target inference host indicates that machine learning services are supported (e.g., in a handover response communication received by the target network node, as described above in conjunction with reference numeral 920), the handover confirmation communication sent to the source network node may indicate that the handover process is approved.

[0200] Additionally, the handover confirmation communication may indicate machine learning inference information associated with the target inference host. For example, the handover confirmation communication sent to the source network node may include an indication of a machine learning model associated with the machine learning service executed on the target inference host, a machine learning inference host identifier associated with the target inference host, and / or machine learning inference data information, etc.

[0201] As indicated by reference numeral 924, the core network node may modify an ML data session associated with a second UE (e.g., a PDU session associated with a second UE that sends machine learning data to a source inference host). For example, the core network node may modify an ML data session associated with a local UPF of a source network node and / or source inference host. For example, a core network node (e.g., an AMF and / or an SMF) may modify a data path associated with an ML data session from a source network node and / or source inference host (e.g., an ML data PDU session established with a local UPF associated with the source network node and / or source inference host) to a target network node and / or target inference host (e.g., an ML data PDU session to be established with a local UPF associated with the target network node and / or target inference host). In some aspects, the core network node may send an indication of the modification of the ML data PDU session to a local UPF associated with the source network node and / or source inference host. Additionally or alternatively, the core network node may send an indication of the establishment of the ML data PDU session to a local UPF associated with the target network node and / or target inference host. For example, the AMF may send a PDU session update request (e.g., including a PDU session identifier associated with the machine learning data service, an indication of the source network node and / or the source inference host and the associated local UPF, an indication of the target network node identifier and / or the target inference host identifier and the associated local UPF, a UE ML context, etc.) to the SMF based at least in part on a handover required message received from the source network node (e.g., as described above in conjunction with reference number 912) and a target network node selection and / or a target inference host selection (e.g., as described above in conjunction with reference number 914). Based at least in part on the received PDU session update request message, the SMF may transmit a PDU session modification request to a local UPF associated with the source network node and / or the source inference host, and the SMF may transmit a PDU session modification request to a local UPF associated with the target network node and / or the target inference host. After receiving a PDU session modification response from a local UPF associated with a source network node and / or a source inference host and receiving a PDU session modification response from a local UPF associated with a target network node and / or a target inference host, the SMF may respond to the AMF with a PDU session update request response communication.

[0202] As shown by reference numeral 926, a core network node (e.g., an AMF) may send a handover command communication, and a source network node may receive the handover command communication. In some aspects, the handover command communication may be sent via an N2 interface. The handover command communication may indicate whether the handover process has been accepted by the target network node. Additionally, the handover command communication may indicate an identifier associated with the target network node, an identifier associated with the target inference host, and / or machine learning data session information (e.g., one or more machine learning data session identifiers and related machine learning data session parameters (such as machine learning data session type)). Additionally, the handover command communication received by the source network node (e.g., from the core network node) may indicate machine learning inference information associated with the target inference host. For example, a handover command communication sent to the source network node may include an indication of a machine learning model associated with a machine learning service executed on the target inference host and / or machine learning data (e.g., machine learning data service type, name, identifier, and / or index), and the like.

[0203] As indicated by reference numeral 928, the source network node may send a handover confirmation communication based at least in part on the source network node receiving the handover command communication from the core network node, and the source inference host may receive the handover confirmation communication. For example, the source network node may indicate to the source inference host that the handover of the first UE and / or the second UE is confirmed and / or in progress.

[0204] In some aspects, as shown by reference numeral 932, a source inference host may send (e.g., via a local UPF) machine learning data and / or machine learning inferences associated with a machine learning service for a first UE and / or a second UE, and a target inference host may receive the machine learning data and / or the machine learning inferences. For example, the source inference host may send machine learning data (e.g., provided by the second UE) and / or machine learning inferences (e.g., generated by the source inference host) to a local UPF associated with a source network node and / or a source inference host. The local UPF associated with the source network node and / or the source inference host may send machine learning data and / or machine learning inferences to a local UPF associated with a target network node and / or a target inference host. The local UPF associated with the target network node and / or the source inference host may send machine learning data and / or machine learning inferences to the target inference host.

[0205] For example, based at least in part on receiving a handover confirmation communication from a source network node, a source inference host may send machine learning data and / or machine learning inferences associated with a machine learning service for a first UE and / or a second UE. The machine learning data may be buffered by the source inference host (e.g., machine learning data provided by the second UE). The machine learning inference may be generated by the source inference host (e.g., using machine learning data provided by the second UE). This may enable a target inference host to collect machine learning data that will be used to train and / or aggregate a machine learning model associated with the target inference host. For example, the target inference host may use machine learning data provided by the second UE and / or machine learning inferences generated by the source inference host to train and / or aggregate a machine learning model that will be used to support a machine learning service for the first UE and / or the second UE. This may reduce the amount of time and / or processing overhead associated with training a machine learning model, among other things.

[0206] In some aspects, a second UE (e.g., a UE that collects and / or provides machine learning data) can be handed over to a target network node before a first UE (e.g., an end-user UE). This can enable the second UE to provide machine learning data to a target inference host so that the target inference host can use the machine learning data used to hand over the first UE to make one or more predictions (e.g., beam or beam pair predictions, etc.). For example, Fig. 9B As shown in reference number 932, the source network node may send a handover command communication indicating that the second UE will perform a handover with the target network node, and the second UE may receive the handover command communication. The handover command communication may be an RRC communication. The handover command may indicate the target network node (e.g., an identifier associated with the target network node), the target inference host (e.g., an identifier associated with the target inference host), machine learning data session information (e.g., one or more machine learning data session identifiers and related machine learning data session parameters, such as a machine learning data session type), machine learning data service information (e.g., a machine learning service type, name, identifier, and / or index as described above in conjunction with reference number 908), and / or other communication parameters to be used by the second UE to establish a connection with the target network node and / or the target inference host.

[0207] The handover command communication may indicate machine learning inference information associated with a target inference host associated with the target network node. For example, the handover command communication may indicate a machine learning model associated with a machine learning service executed on the target inference host, a machine learning inference host identifier associated with the target inference host, and / or machine learning data session information, etc. As part of the handover process, this may enable the second UE to identify the machine learning information associated with the target network node. For example, the second UE may identify an identifier of the target inference host, a machine learning model associated with the target inference host, and / or machine learning data session information associated with the target inference host.

[0208] As indicated by reference numeral 934, the second UE may send a communication for establishing a communication connection (e.g., an RRC connection) with the target network node, and the target network node may receive the communication. For example, the second UE may send a RACH communication (e.g., msgA or msg1) or an RRC connection communication. The communication may indicate machine learning data session information associated with a machine learning data service provided by the second UE (e.g., a data session identifier and associated QoS parameters), information associated with the machine learning data (e.g., type or format or structure), information associated with the machine learning data service (e.g., information associated with a machine learning model associated with the machine learning service executed on the second UE, an indication of one or more machine learning service types associated with a QoS profile or QoS flow or 5QI, one or more machine learning model identifiers associated with the machine learning service, and / or one or more machine learning inference host identifiers associated with the machine learning service), and the like. For example, as part of establishing an RRC connection with the target network node, the communication may indicate the machine learning data service information. In a manner similar to that described in conjunction with reference numeral 858, an ML data session (e.g., one or more PDU sessions) may be established between the second UE and the target inference host.

[0209] As indicated by reference numeral 938, the second UE may send second machine learning data for a machine learning service (e.g., based at least in part on receiving a handover command communication to handover the second UE to the target network node and / or based at least in part on establishing an ML data session), and the target inference host may receive the second machine learning data. For example, the second UE may process one or more sensor measurements via a machine learning model executed on the second UE to obtain second machine learning data associated with the machine learning service. The second UE may send the second machine learning data to the second network node or the second inference host via the ML data session. This may enable the target inference host to collect machine learning data that may be used to perform machine learning operations associated with the second UE and / or the first UE (e.g., before the first UE is handed over to the target network node).

[0210] For example, the target inference host may process machine learning data (e.g., received via an ML data session or received during a handover process, as described in more detail elsewhere herein) to train and / or converge a machine learning model executed on the target inference host. The target inference host may process the machine learning data provided by the second UE to perform one or more machine learning operations associated with the first UE. For example, the target inference host may generate one or more inferences or predictions associated with communication parameters to be used by the first UE to communicate with the target network node. For example, the target inference host may use the machine learning model to determine an optimal beam, an optimal beam pair, an optimal time / frequency resource, an optimal transmit power, and / or an optimal MCS, etc., to be used by the first UE to communicate with the target network node.

[0211] As shown by reference numeral 940, the target inference host may send one or more machine learning predictions associated with the first UE (e.g., where the one or more machine learning predictions are based at least in part on machine learning data generated by a second UE and provided to the target inference host), and the target network node may receive the one or more machine learning predictions. As shown by reference numeral 942, the target network node may send an indication of one or more machine learning predictions associated with the first UE, and the source network node may receive the indication. For example, the target network node may send the indication of the one or more machine learning predictions via an Xn interface. The source network node and the target network node may continue to perform a handover procedure for the first UE based on the one or more predictions (e.g., by configuring the first UE to communicate with the target network node using one or more communication parameters or beams indicated by the one or more predictions).

[0212] In some aspects, the source network node may send UE context information and / or data associated with the first UE, and the target network node may receive UE context information and / or data associated with the first UE, as indicated by reference numeral 944. For example, the UE context information may include an identifier associated with the first UE and / or other information associated with the first UE (e.g., machine learning capabilities or machine learning information) to be stored by the target network node.

[0213] As indicated at reference numeral 946, the source network node may send a handover command communication, and the first UE may receive the handover command communication. For example, the source network node may send an RRC configuration message to the first UE (e.g., in an RRC reconfiguration message) by forwarding an RRC configuration associated with the target network node to the first UE. For example, the target network node may indicate an RRC configuration to be used by the first UE in the handover command communication (e.g., sent as described above in connection with reference numeral 926). For example, the RRC configuration indicated in the handover command communication may be based at least in part on one or more predictions generated by the target inference host (e.g., sent as described above in connection with reference numeral 942).

[0214] The first UE may change the RRC connection from the source network node to the target network node based at least in part on the handover command. As shown by reference numeral 948, the first UE may send a RACH or RRC connection message to the target network node. The RRC connection message may indicate that the first UE has changed the RRC connection from the source network node to the target network node. In some aspects, the target network node may send a UE context release message to the source network node. The UE context release message may indicate that the handover of the first UE to the target network node is successful.

[0215] The first UE may communicate with the target network node based at least in part on performing the handover procedure. For example, the first UE may communicate with the target network node using one or more communication parameters or beams indicated by one or more predictions generated by the target inference host. Thus, machine learning services associated with the first UE may continue immediately when the first UE is handed over to the target network node (e.g., without delays associated with establishing an ML data session for the second UE to provide machine learning data to the target inference host or training a machine learning model executed on the target inference host).

[0216] Thus, when the first UE is handed over from the source network node to the target network node, machine learning service continuity and machine learning data continuity can be maintained. This enables the target network node to perform a handover process using one or more machine learning predictions associated with the first UE. Additionally, the handover process described herein can increase the likelihood that the first UE is handed over to a network node capable of performing machine learning operations associated with the machine learning service. By ensuring machine learning service continuity and machine learning data continuity as part of the handover process, communication performance of the first UE (e.g., an end user UE) can be improved (e.g., because the target network node can quickly utilize accurate machine learning predictions to select communication parameters associated with the first UE).

[0217] As indicated above, Fig. 9A and Fig. 9B are provided as examples. Other examples can be found in the Fig. 9A and Fig. 9B Different than described.

[0218] Fig.10 1 is a diagram illustrating an example 1000 associated with UE machine learning service continuity according to the present disclosure. Fig.10 As shown, example 1000 includes a terminal UE (eg, UE 120 or Figure 8 , Fig. 9A and Fig. 9B ), a vehicle UE (e.g., UE 120 or Figure 8 , Fig. 9A and Fig. 9B ), a first network node (NN 1) (eg, network node 110 or Figure 8 , Fig. 9A and Fig. 9B ), a first inference host (e.g., a near RT RIC 325, a non-RT RIC 315, a model inference host 404, an ML inference host 514, or Figure 8 , Fig. 9A and Fig. 9B , etc.), a second network node (NN 2) (e.g., network node 110 or Figure 8 , Fig. 9A and Fig. 9B ), a second inference host (e.g., a near-RT RIC 325, a non-RT RIC 315, a model inference host 404, an ML inference host 514, or Figure 8 , Fig. 9A and Fig. 9B Communication between the target inference host, etc. depicted in FIG.

[0219] like Fig.10 As shown, the handover process for the terminal UE may involve a communication handover (e.g., an RRC connection), a machine learning data handover (e.g., to enable the vehicle UE to provide machine learning data to the second inference host), and / or a machine learning service handover (e.g., to enable the second inference host to provide machine learning services associated with the terminal UE). For example, machine learning data continuity may be achieved by performing a handover process associated with switching the vehicle UE from the first network node to the second network node. Additionally, an ML data session (e.g., a PDU session) may be handed over to the second inference host or established with the second inference host (e.g., as described elsewhere herein (such as in conjunction with Figure 8 , Fig. 9A and Fig. 9B ) described in more detail). Thus, the vehicle UE can be enabled to communicate with the second network node and provide machine learning data (e.g., obtained from sensor measurements or sensor data) to the second inference host.

[0220] Additionally, machine learning service continuity can be achieved by providing machine learning information and / or data associated with the machine learning service to the second inference host (e.g., as described elsewhere herein (such as in conjunction with Figure 8 , Fig. 9A and Fig. 9B ) described in more detail). For example, the second inference host may receive machine learning data, machine learning context, and / or other information to enable the second inference host to quickly train or converge a machine learning model associated with the machine learning service. This may enable the second inference host to perform machine learning operations and / or provide predictions associated with the terminal UE (e.g., using machine learning data provided by the vehicle UE). Communication continuity may be maintained by performing a handover process from the first network node to the second network node associated with the terminal UE. Because machine learning data continuity and / or machine learning service continuity for the terminal UE have been maintained, communications associated with the second network node may be improved (e.g., after the handover process is completed). For example, the terminal UE may be instructed to use one or more communication parameters or beams selected based at least in part on predictions obtained by the second inference host or machine learning operations performed by the second inference host.

[0221] As indicated above, Fig.10 are provided as examples. Other examples can be found in the Fig.10 Different than described.

[0222] Fig.11 1 is a diagram illustrating an example process 1100 performed, for example, by a UE according to the present disclosure. The example process 1100 is a diagram in which a UE (e.g., UE 120, a vehicle UE, or a combination thereof) performs a Figure 8 , Fig. 9A and Fig. 9B An example of a second UE) performing operations associated with UE machine learning service continuity is depicted and described.

[0223] like Fig.11 As shown, in some aspects, process 1100 may include sending first machine learning data associated with a machine learning service to a first network node for use by a first inference host associated with the first network node (block 1110). Fig.15 The communication manager 140 and / or sending component 1504 depicted in the figure can send first machine learning data associated with the machine learning service to the first network node for use by a first inference host associated with the first network node, as described above.

[0224] like Fig.11 As further shown, in some aspects, process 1100 may include receiving a handover command communication from a first network node, the handover command communication indicating that the UE is to perform a handover from the first network node to a second network node, wherein the handover command communication indicates machine learning inference information associated with a second inference host associated with the second network node (block 1120). For example, the UE (e.g., using Fig.15 The communication manager 140 and / or receiving component 1502 depicted in the figure can receive a handover command communication from the first network node, the handover command communication indicating that the UE will perform a handover from the first network node to the second network node, wherein the handover command communication indicates machine learning inference information associated with a second inference host associated with the second network node, as described above.

[0225] like Fig.11 As further shown, in some aspects, process 1100 may include, based at least in part on receiving the handover command communication, sending second machine learning data associated with the machine learning service to the second network node for use by a second inference host associated with the second network node (block 1130). Fig.15 The communication manager 140 and / or sending component 1504 depicted in the figure can send second machine learning data associated with the machine learning service to the second network node for use by a second inference host associated with the second network node based at least in part on receiving the handover command communication, as described above.

[0226] Process 1100 may include additional aspects, such as any single aspect or any combination of aspects described below and / or in conjunction with one or more other processes described elsewhere herein.

[0227] In a first aspect, a UE is associated with a vehicle.

[0228] In a second aspect, alone or in combination with the first aspect, process 1100 includes: obtaining one or more sensor measurements associated with one or more objects via one or more sensors; and processing the one or more sensor measurements via a machine learning model executed on the UE to obtain machine learning data associated with a machine learning service.

[0229] In a third aspect, either alone or in combination with one or more of the first and second aspects, at least one of the first machine learning data or the second machine learning data comprises at least one of: the geographic location of the UE, the direction of travel of the UE, the speed of the UE, beam information associated with another UE (e.g., an end user UE) co-located with the UE, or data extracted from one or more sensor measurements.

[0230] In a fourth aspect, either alone or in combination with one or more of the first to third aspects, process 1100 includes sending a RACH connection communication or an RRC connection communication to a second network node indicating at least one of the following: information associated with machine learning data, information associated with one or more machine learning models for generating machine learning data, information associated with a machine learning service, an indication of one or more machine learning types associated with the machine learning service, one or more machine learning model identifiers associated with the machine learning service, or one or more machine learning inference host identifiers associated with the machine learning service.

[0231] In a fifth aspect, alone or in combination with one or more of aspects 1 to 4, process 1100 includes: receiving a first PDU session communication from a second network node, the first PDU session communication indicating PDU information associated with a machine learning service; and in response to the first PDU session communication, sending a second PDU session communication to the second network node to establish a PDU session with a second inference host.

[0232] In a sixth aspect, either alone or in combination with one or more of aspects 1 to 5, sending second machine learning data comprises: processing one or more sensor measurements via a machine learning model executed on the UE to obtain second machine learning data associated with a machine learning service; and sending the second machine learning data to a second network node or a second inference host via a PDU session.

[0233] In a seventh aspect, either alone or in combination with one or more of the first to sixth aspects, the UE is associated with another UE, and wherein the machine learning data associated with the machine learning service is associated with the other UE.

[0234] In an eighth aspect, either alone or in combination with one or more of the first to seventh aspects, the machine learning service comprises at least one of beam prediction, handover prediction or beam blocking prediction.

[0235] although Fig.11 An example block diagram of process 1100 is shown, but in some aspects, process 1100 may include Fig.11 Additional blocks, fewer blocks, different blocks, or blocks arranged in a different manner than those depicted in the process 1100. Additionally or alternatively, two or more blocks of the blocks of process 1100 may be performed in parallel.

[0236] Fig.12 1 is a diagram illustrating an example process 1200 performed, for example, by a first network node according to the present disclosure. Example process 1200 is an example in which a first network node (e.g., network node 110 or a source network node) performs operations associated with UE machine learning service continuity.

[0237] like Fig.12 As shown, in some aspects, process 1200 may include receiving a first handover communication from a first inference host associated with a first network node, the first handover communication indicating first machine learning inference information associated with a machine learning service, the machine learning service associated with machine learning data associated with a first UE (block 1210). For example, the first network node (e.g., using Fig.16 The communication manager 150 and / or receiving component 1602 depicted in the figure can receive a first handover communication from a first inference host associated with the first network node, the first handover communication indicating first machine learning inference information associated with a machine learning service, which machine learning service is associated with machine learning data associated with the first UE, as described above.

[0238] like Fig.12 As further shown, in some aspects, process 1200 may include sending a handover request communication associated with a handover procedure for the first UE to a second network node or a core network node, wherein the first network node selects the second network node for the handover procedure based at least in part on the first machine learning inference information and a machine learning service associated with the second network node (block 1220). For example, the first network node (e.g., using Fig.16 The communication manager 150 and / or sending component 1604 depicted in the figure can send a handover request communication associated with a handover process for the first UE to a second network node or a core network node, wherein the first network node selects the second network node for the handover process based at least in part on the first machine learning inference information and the machine learning service associated with the second network node, as described above.

[0239] like Fig.12 As further shown, in some aspects, process 1200 may include receiving a handover response communication associated with the handover process from a second network node or a core network node, the handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node (block 1230). Fig.16 The communication manager 150 and / or receiving component 1602 depicted in the figure can receive a handover response communication associated with the handover process from a second network node or a core network node, the handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node, as described above.

[0240] Process 1200 may include additional aspects, such as any single aspect or any combination of aspects described below and / or in conjunction with one or more other processes described elsewhere herein.

[0241] In a first aspect, process 1200 includes: receiving a measurement report from a first UE or a second UE, the measurement report indicating one or more measurements associated with one or more candidate network nodes including a second network node; and determining to hand over at least one of the first UE or the second UE to the second network node based at least in part on the measurement report, first machine learning inference information, and a machine learning service associated with the second network node.

[0242] In a second aspect, alone or in combination with the first aspect, process 1200 includes sending a second handover communication indicating at least one of machine learning service information or machine learning data to the first inference host, wherein the first handover communication is responsive to the second handover communication.

[0243] In a third aspect, either alone or in combination with one or more of the first and second aspects, the machine learning data comprises at least one of: an identification of a machine learning data service provided by a first UE, an identifier associated with the first UE, an indication of one or more machine learning models associated with the machine learning data service executed on the first UE, or an indication of one or more machine learning data types associated with one or more machine learning operations performed by the first UE on the machine learning data service.

[0244] In a fourth aspect, either alone or in combination with one or more of aspects 1 to 3, the first machine learning inference information indicates at least one of: one or more machine learning models associated with the machine learning service executed on the first inference host, or information associated with the machine learning service or a state of machine learning data associated with the machine learning service.

[0245] In a fifth aspect, alone or in combination with one or more of the first to fourth aspects, the handover request communication indicates first machine learning inference information received from a first inference host.

[0246] In a sixth aspect, alone or in combination with one or more of the first to fifth aspects, the handover request communication is sent via an Xn interface.

[0247] In a seventh aspect, either alone or in combination with one or more of aspects 1 to 6, process 1200 comprises sending a third handover communication to the first inference host based at least in part on receiving a handover response communication, the handover response communication indicating that handover of at least one of the first UE or the second UE to the second network node is confirmed.

[0248] In an eighth aspect, either alone or in combination with one or more of aspects 1 to 7, process 1200 includes sending an RRC communication to a first UE indicating a handover command for the first UE, wherein the RRC communication indicates second machine learning inference information associated with a second inference host.

[0249] In a ninth aspect, alone or in combination with one or more of aspects 1 to 8, process 1200 includes: after sending a handover request communication, receiving machine learning data for a machine learning service associated with a first UE from a first inference host; sending machine learning data for a machine learning service associated with the first UE to a second network node or a core network node; and sending context information associated with the first UE to the second network node or the core network node, the context information including at least one of machine learning service information or machine learning data.

[0250] In a tenth aspect, alone or in combination with one or more of aspects one to nine, process 1200 includes: sending UE context information and data associated with a second UE co-located with a first UE to a second network node; receiving one or more machine learning predictions associated with the second UE from the second network node, wherein the one or more machine learning predictions are at least partially based on machine learning data generated by the first UE; and sending an RRC communication indicating a handover command to the second UE, wherein the RRC communication indicates that the second UE will connect to the second network node based on the one or more machine learning predictions.

[0251] In the eleventh aspect, alone or in combination with one or more of the first to tenth aspects, the core network node is associated with the AMF.

[0252] although Fig.12 An example block diagram of process 1200 is shown, but in some aspects, process 1200 may include Fig.12 Additional blocks, fewer blocks, different blocks, or blocks arranged in a different manner than those depicted in the process 1200. Additionally or alternatively, two or more blocks of the blocks of process 1200 may be performed in parallel.

[0253] Fig.13 1 is a diagram illustrating an example process 1300 performed, for example, by a first network node according to the present disclosure. Example process 1300 is an example in which a first network node (e.g., network node 110 and / or a target network node) performs operations associated with UE machine learning service continuity.

[0254] like Fig.13 As shown, in some aspects, process 1300 may include receiving a handover request communication associated with a handover procedure for a UE from a second network node or a core network node, wherein the handover request communication indicates first machine learning inference information associated with a machine learning service and machine learning data associated with the UE (block 1310). For example, a first network node (e.g., using Fig.16 The communication manager 150 and / or receiving component 1602 depicted in the figure can receive a handover request communication associated with a handover process for the UE from a second network node or a core network node, wherein the handover request communication indicates first machine learning inference information associated with a machine learning service and machine learning data associated with the UE, as described above.

[0255] like Fig.13 As further shown, in some aspects, process 1300 may include communicating with a first inference host associated with a first network node to determine that the first inference host is capable of supporting machine learning services (block 1320). Fig.16 The communication manager 150, sending component 1604 and / or receiving component 1602 depicted in can communicate with a first inference host associated with a first network node to determine that the first inference host is capable of supporting machine learning services, as described above.

[0256] like Fig.13 As further shown, in some aspects, process 1300 may include sending a handover response communication indicating second machine learning inference information associated with the first inference host to a second network node or a core network node based at least in part on the first inference host being able to support the machine learning service (block 1330). Fig.16 The communication manager 150 and / or sending component 1604 depicted in the figure can send a handover response communication indicating second machine learning inference information associated with the first inference host to the second network node or core network node based at least in part on the first inference host being able to support the machine learning service, as described above.

[0257] Process 1300 may include additional aspects, such as any single aspect or any combination of aspects described below and / or in conjunction with one or more other processes described elsewhere herein.

[0258] In a first aspect, communicating to determine that a first inference host is capable of supporting a machine learning service includes: sending a first handover communication indicating at least one of first machine learning inference information or machine learning data to the first inference host; and receiving a second handover communication from the first inference host indicating whether the first inference host is capable of supporting the machine learning service, wherein the second handover communication indicates second machine learning inference information associated with the first inference host.

[0259] In a second aspect, either alone or in combination with the first aspect, the first machine learning inference information indicates at least one of: an identifier of a second inference host associated with the machine learning service and the second network node, a machine learning model executed on the second inference host, or other machine learning data associated with the machine learning service associated with the second inference host.

[0260] In a third aspect, either alone or in combination with one or more of the first and second aspects, the second machine learning inference information indicates at least one of: an identifier of the first inference host, a machine learning model executed on the first inference host, or machine learning data associated with a machine learning service associated with the first inference host.

[0261] In a fourth aspect, alone or in combination with one or more of the first to third aspects, process 1300 includes receiving a RACH connection communication or an RRC connection communication from a UE indicating machine learning service information associated with the UE, the machine learning service information indicating at least one of the following: machine learning data associated with the machine learning service, machine learning type or service information associated with the machine learning service, or information associated with a machine learning model associated with the machine learning service executed on the UE.

[0262] In a fifth aspect, either alone or in combination with one or more of aspects 1 to 4, process 1300 includes receiving UE context information from a second network node or a core network node, the UE context information comprising at least one of machine learning service information or machine learning data.

[0263] In a sixth aspect, alone or in combination with one or more of aspects 1 to 5, process 1300 includes: sending a first PDU communication to a first inference host, the first PDU communication indicating first machine learning inference information associated with a machine learning service for the UE; and receiving a second PDU communication from the first inference host, the second PDU communication indicating PDU session information of a PDU session associated with the machine learning service.

[0264] In a seventh aspect, either alone or in combination with one or more of aspects 1 to 6, process 1300 includes sending an indication of PDU session information for a PDU session associated with a machine learning service to a UE.

[0265] In an eighth aspect, either alone or in combination with one or more of aspects 1 to 7, process 1300 includes sending a PDU establishment communication to a core network node, the PDU establishment communication indicating first machine learning inference information associated with a machine learning service for the UE, so that a PDU session associated with the machine learning service is associated with the first inference host.

[0266] In a ninth aspect, either alone or in combination with one or more of aspects 1 to 8, process 1300 includes: after sending a handover response communication, receiving machine learning data for a machine learning service associated with the UE from a second network node; and sending the machine learning data to a first inference host.

[0267] In a tenth aspect, alone or in combination with one or more of aspects one to nine, process 1300 includes: receiving one or more machine learning predictions associated with another UE from a first inference host, the one or more machine learning predictions being based at least in part on machine learning data provided by the UE; and sending an indication of the one or more machine learning predictions to a second network node to facilitate handing over the other UE to the first network node based on the one or more machine learning predictions.

[0268] In an eleventh aspect, alone or in combination with one or more of the first to tenth aspects, process 1300 includes receiving an RRC connection request communication from another UE based on one or more machine learning predictions.

[0269] In a twelfth aspect, either alone or in combination with one or more of the first to eleventh aspects, the one or more machine learning predictions include a beam or beam pair to be used by another UE to establish a connection with the first network node.

[0270] although Fig.13 Example blocks of process 1300 are shown, but in some aspects, process 1300 may include Fig.13 Additional blocks, fewer blocks, different blocks, or blocks arranged in a different manner than those depicted in the process 1300. Additionally or alternatively, two or more blocks of the blocks of process 1300 may be performed in parallel.

[0271] Fig.14 1 is a diagram illustrating an example process 1400 performed, for example, by a core network node according to the present disclosure. Example process 1400 is an example in which a core network node (e.g., core network node 160, AMF, SMF and / or UPF, etc.) performs operations associated with UE machine learning service continuity.

[0272] like Fig.14 As shown, in some aspects, process 1400 may include receiving a first handover request communication associated with a handover procedure for a UE from a first network node, wherein the first handover request communication indicates at least one of first machine learning inference information associated with a machine learning service, machine learning data associated with the UE, or a first inference host associated with the first network node (block 1410). For example, a core network node (e.g., using Fig.17 The communication manager 170 and / or receiving component 1702 depicted in the figure can receive a first handover request communication associated with a handover process for a UE from a first network node, wherein the first handover request communication indicates at least one of first machine learning inference information associated with a machine learning service, machine learning data associated with the UE, or a first inference host associated with the first network node, as described above.

[0273] like Fig.14 As further shown, in some aspects, process 1400 may include sending a second handover request communication associated with a handover procedure for the UE to a second network node, wherein the second network node is selected (e.g., by a core network node) for the handover procedure based at least in part on the first machine learning inference information and the machine learning capability associated with the second network node (block 1420). Fig.17 The communication manager 170 and / or sending component 1704 depicted in the figure can send a second handover request communication associated with a handover process for the UE to a second network node, wherein the second network node is selected for the handover process based at least in part on the first machine learning inference information and the machine learning capabilities associated with the second network node, as described above.

[0274] like Fig.14As further shown, in some aspects, process 1400 may include receiving a first handover response communication associated with the handover process from a second network node, the first handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node (block 1430). Fig.17 The communication manager 170 and / or receiving component 1702 depicted in the figure can receive a first handover response communication associated with the handover process from the second network node, and the first handover response communication indicates second machine learning inference information associated with a second inference host associated with the second network node, as described above.

[0275] Process 1400 may include additional aspects, such as any single aspect or any combination of aspects described below and / or in conjunction with one or more other processes described elsewhere herein.

[0276] In a first aspect, process 1400 includes selecting a second network node for a handover process based at least in part on at least one of first machine learning inference information, machine learning capabilities associated with the second network node, or capabilities associated with a second inference host.

[0277] In a second aspect, alone or in combination with the first aspect, process 1400 includes sending a second handover response communication associated with the handover process to the first network node, the second handover response communication indicating second machine learning inference information.

[0278] In a third aspect, either alone or in combination with one or more of the first and second aspects, the first machine learning inference information includes at least one of: an identifier of a first inference host, an indication of a machine learning model associated with a machine learning service executed on the first inference host, or machine learning data associated with the machine learning service.

[0279] In a fourth aspect, either alone or in combination with one or more of aspects 1 to 3, the machine learning data indicates at least one of: an identifier of a machine learning data service associated with the UE, a machine learning model associated with the machine learning service executed on the UE, or a machine learning data type or format associated with the machine learning service.

[0280] In a fifth aspect, either alone or in combination with one or more of aspects 1 to 4, the second machine learning inference information indicates at least one of: an identifier of a second inference host, a machine learning model associated with the machine learning service executed on the second inference host, or machine learning data associated with the machine learning service.

[0281] In a sixth aspect, alone or in combination with one or more of the first to fifth aspects, the first handover request communication is received via an N2 interface.

[0282] In a seventh aspect, either alone or in combination with one or more of the first to sixth aspects, the core network node is associated with an AMF.

[0283] although Fig.14 Example blocks of process 1400 are shown, but in some aspects, process 1400 may include Fig.14 Additional blocks, fewer blocks, different blocks, or blocks arranged in a different manner than those depicted in the process 1400. Additionally or alternatively, two or more blocks of the blocks of process 1400 may be performed in parallel.

[0284] Fig.15 1 is a diagram of an example apparatus 1500 for wireless communication. Apparatus 1500 may be a UE, or a UE may include apparatus 1500. In some aspects, apparatus 1500 includes a receiving component 1502 and a sending component 1504, which may communicate with each other (e.g., via one or more buses and / or one or more other components). As shown, apparatus 1500 may use receiving component 1502 and sending component 1504 to communicate with another apparatus 1506 (such as a UE, a base station, a network node, a network entity, a core network node, or another wireless communication device). As further shown, apparatus 1500 may include a communication manager 140. Communication manager 140 may include one or more of a sensor measurement component 1508 and / or an ML processing component 1510, etc.

[0285] In some aspects, the apparatus 1500 may be configured to perform Figure 8 , Fig. 9A , Fig. 9B and Fig.10 Additionally or alternatively, the apparatus 1500 may be configured to perform one or more of the processes described herein (such as Fig.11 In some aspects, Fig.15 The device 1500 and / or one or more components shown may include a combination of Figure 2 Additionally or alternatively, Fig.15 One or more of the components shown may be combined with Figure 2Additionally or alternatively, one or more components in the component set may be implemented at least in part as software stored in a memory. For example, a component (or a portion of a component) may be implemented as instructions or code stored in a non-transitory computer-readable medium and can be executed by a controller or processor to perform the function or operation of the component.

[0286] The receiving component 1502 may receive communications, such as reference signals, control information, data communications, or combinations thereof, from the device 1506. The receiving component 1502 may provide the received communications to one or more other components of the device 1500. In some aspects, the receiving component 1502 may perform signal processing (such as filtering, amplification, demodulation, analog-to-digital conversion, demultiplexing, deinterleaving, demapping, equalization, interference cancellation, or decoding, etc.) on the received communications and may provide the processed signals to one or more other components of the device 1500. In some aspects, the receiving component 1502 may include combining Figure 2 One or more antennas, modems, demodulators, MIMO detectors, receive processors, controllers / processors, memories, or combinations thereof of the described UE.

[0287] Transmit component 1504 may transmit communications, such as reference signals, control information, data communications, or combinations thereof, to device 1506. In some aspects, one or more other components of device 1500 may generate communications and may provide the generated communications to transmit component 1504 for transmission to device 1506. In some aspects, transmit component 1504 may perform signal processing (such as filtering, amplification, modulation, digital-to-analog conversion, multiplexing, interleaving, mapping, or encoding, etc.) on the generated communications and may transmit the processed signals to device 1506. In some aspects, transmit component 1504 may include combining Figure 2 One or more antennas, modems, modulators, transmit MIMO processors, transmit processors, controllers / processors, memories, or combinations thereof of the described UE. In some aspects, the transmit component 1504 can be co-located with the receive component 1502 in a transceiver.

[0288] The sending component 1504 can send first machine learning data for machine learning services to a first inference host associated with the first network node. The receiving component 1502 can receive a handover command communication from the first network node, the handover command communication indicating that the UE will perform a handover with the second network node, wherein the handover command communication indicates machine learning information associated with a second inference host associated with the second network node. The sending component 1504 can send second machine learning data for machine learning services to a second inference host associated with the second network node based at least in part on receiving the handover command communication.

[0289] The sensor measurement component 1508 can obtain one or more sensor measurements associated with one or more objects via one or more sensors. The ML processing component 1510 can process the one or more sensor measurements via a machine learning model executed on the UE to obtain machine learning data associated with a machine learning service.

[0290] The sending component 1504 can send a RACH connection communication or an RRC connection communication to the second network node indicating at least one of the following: information associated with the machine learning data, information associated with one or more machine learning models used to generate the machine learning data, information associated with the machine learning service, an indication of one or more machine learning types associated with the machine learning service, one or more machine learning model identifiers associated with the machine learning service, or one or more machine learning inference host identifiers associated with the machine learning service.

[0291] The receiving component 1502 can receive a first PDU session communication from a second network node, the first PDU session communication indicating PDU information associated with a machine learning service. The sending component 1504 can send a second PDU session communication to the second network node in response to the first PDU session communication to establish a PDU session with a second inference host.

[0292] Fig.15 The number and arrangement of components shown are provided as examples. In practice, there may be Fig.15 Additional components, fewer components, different components, or components arranged in a different manner than those shown. Fig.15 Two or more components shown may be implemented in a single component, or Fig.15 The single component shown may be implemented as multiple distributed components. Additionally or alternatively, Fig.15 The component set (one or more components) shown may be described as being executable by Fig.15 Another collection of components shown performs one or more functions.

[0293] Fig.161 is a diagram of an example apparatus 1600 for wireless communication. Apparatus 1600 may be a network node, or a network node may include apparatus 1600. In some aspects, apparatus 1600 includes a receiving component 1602 and a sending component 1604, which may communicate with each other (e.g., via one or more buses and / or one or more other components). As shown, apparatus 1600 may use receiving component 1602 and sending component 1604 to communicate with another apparatus 1606 (such as a UE, a base station, a network node, a network entity, a core network node, or another wireless communication device). As further shown, apparatus 1600 may include a communication manager 150. Communication manager 150 may include determining component 1608, etc.

[0294] In some aspects, the apparatus 1600 may be configured to perform Figure 8 , Fig. 9A , Fig. 9B and Fig.10 Additionally or alternatively, the apparatus 1600 may be configured to perform one or more processes described herein, such as Fig.12 The process 1200 Fig.13 In some aspects, Fig.16 The device 1600 and / or one or more components shown may include a combination of Figure 2 Additionally or alternatively, Fig.16 One or more of the components shown may be combined with Figure 2 Additionally or alternatively, one or more components in the component set may be implemented at least in part as software stored in a memory. For example, a component (or a portion of a component) may be implemented as instructions or code stored in a non-transitory computer-readable medium and can be executed by a controller or processor to perform the function or operation of the component.

[0295] The receiving component 1602 may receive communications from the device 1606, such as reference signals, control information, data communications, or a combination thereof. The receiving component 1602 may provide the received communications to one or more other components of the device 1600. In some aspects, the receiving component 1602 may perform signal processing (such as filtering, amplification, demodulation, analog-to-digital conversion, demultiplexing, deinterleaving, demapping, equalization, interference cancellation, or decoding, etc.) on the received communications and may provide the processed signals to one or more other components of the device 1600. In some aspects, the receiving component 1602 may include combining Figure 2 One or more antennas, modems, demodulators, MIMO detectors, receive processors, controllers / processors, memories, or combinations thereof, of the described network nodes.

[0296] Transmit component 1604 may transmit communications, such as reference signals, control information, data communications, or combinations thereof, to device 1606. In some aspects, one or more other components of device 1600 may generate communications and may provide the generated communications to transmit component 1604 for transmission to device 1606. In some aspects, transmit component 1604 may perform signal processing (such as filtering, amplification, modulation, digital-to-analog conversion, multiplexing, interleaving, mapping, or encoding, etc.) on the generated communications and may transmit the processed signals to device 1606. In some aspects, transmit component 1604 may include combining Figure 2 One or more antennas, modems, modulators, transmit MIMO processors, transmit processors, controllers / processors, memories, or combinations thereof of the described network nodes. In some aspects, the transmit component 1604 can be co-located with the receive component 1602 in a transceiver.

[0297] The receiving component 1602 may receive a first handover communication from a first inference host associated with the first network node, the first handover communication indicating first machine learning inference information associated with a machine learning service using machine learning data associated with the first UE. The sending component 1604 may send a first handover request communication associated with a handover process for the first UE to a second network node or a core network node, wherein the second network node is selected for the handover process based at least in part on at least one of the first machine learning inference information or the machine learning service associated with the second network node. The receiving component 1602 may receive a handover response communication associated with the handover process from the second network node or the core network node, the handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node.

[0298] Receiving component 1602 can receive a measurement report from the first UE or the second UE, the measurement report indicating one or more measurements associated with one or more candidate network nodes including the second network node.

[0299] Determining component 1608 may determine to hand over at least one of the first UE or the second UE to the second network node based at least in part on the measurement report, the first machine learning inference information, and a machine learning service associated with the second network node.

[0300] The sending component 1604 can send a second handover communication indicating at least one of the machine learning service information or the machine learning data to the first inference host, wherein the first handover communication is responsive to the second handover communication.

[0301] The sending component 1604 can send a third handover communication to the first inference host based at least in part on receiving the handover response communication indicating that handover of at least one of the first UE or the second UE to the second network node is confirmed.

[0302] The sending component 1604 can send an RRC communication indicating a handover command for the first UE to the first UE, wherein the RRC communication indicates second machine learning inference information associated with the second inference host.

[0303] The receiving component 1602 can receive machine learning data for a machine learning service associated with the first UE from the first inference host after sending the handover request communication.

[0304] The sending component 1604 can send machine learning data associated with the first UE for the machine learning service to the second network node or the core network node.

[0305] The sending component 1604 can send context information associated with the first UE to the second network node or the core network node, and the context information includes at least one of machine learning service information or machine learning data.

[0306] Transmitting component 1604 may transmit UE context information and data associated with the second UE to the second network node.

[0307] Receiving component 1602 may receive one or more machine learning predictions associated with a second UE from a second network node, wherein the one or more machine learning predictions are based at least in part on machine learning data generated by the first UE. Sending component 1604 may send an RRC communication indicating a handover command to the second UE, wherein the RRC communication indicates that the second UE will connect with the second network node based on the one or more machine learning predictions.

[0308] The receiving component 1602 may receive a handover request communication associated with a handover procedure for the UE from a second network node or a core network node, wherein the handover request communication indicates at least one of first machine learning inference information associated with a machine learning service or machine learning data associated with the UE. The receiving component 1602 and / or the sending component 1608 may communicate with a first inference host associated with the first network node to determine that the first inference host is capable of supporting the machine learning service. The sending component 1604 may send a handover response communication indicating second machine learning inference information associated with the first inference host to the second network node or the core network node based at least in part on the first inference host being capable of supporting the machine learning service.

[0309] The receiving component 1602 can receive a RACH connection communication or an RRC connection communication from the UE indicating machine learning service information associated with the UE, wherein the machine learning service information indicates at least one of the following: machine learning data associated with the machine learning service, machine learning type or service information associated with the machine learning service, or information associated with a machine learning model associated with the machine learning service executed on the UE.

[0310] The receiving component 1602 can receive UE context information from the second network node or the core network node, and the UE context information includes at least one of machine learning service information or machine learning data.

[0311] The sending component 1604 can send a first PDU communication to the first inference host, the first PDU communication indicating first machine learning inference information associated with the machine learning service for the UE. The receiving component 1602 can receive a second PDU communication from the first inference host, the second PDU communication indicating PDU session information of a PDU session associated with the machine learning service.

[0312] The sending component 1604 may send an indication of PDU session information for a PDU session associated with the machine learning service to the UE. The sending component 1604 may send a PDU setup communication to the core network node, the PDU setup communication indicating first machine learning inference information associated with the machine learning service for the UE to associate the PDU session associated with the machine learning service with the first inference host.

[0313] Receiving component 1602 may receive, from the second network node, machine learning data associated with the UE for a machine learning service after sending the handover response communication.

[0314] The sending component 1604 can send the machine learning data to the first inference host. The receiving component 1602 can receive one or more machine learning predictions associated with another UE from the first inference host, the one or more machine learning predictions based at least in part on the machine learning data provided by the UE.

[0315] The sending component 1604 can send an indication of the one or more machine learning predictions to the second network node to facilitate handing over another UE to the first network node based on the one or more machine learning predictions.

[0316] Receiving component 1602 can receive an RRC connection request communication from another UE based on one or more machine learning predictions.

[0317] Fig.16 The number and arrangement of components shown are provided as examples. In practice, there may be Fig.16Additional components, fewer components, different components, or components arranged in a different manner than those shown. Fig.16 Two or more components shown may be implemented in a single component, or Fig.16 The single component shown may be implemented as multiple distributed components. Additionally or alternatively, Fig.16 The component set (one or more components) shown may be described as being executable by Fig.16 Another collection of components shown performs one or more functions.

[0318] Fig.17 1 is a diagram of an example apparatus 1700 for wireless communication. Apparatus 1700 may be a core network node, or a core network node may include apparatus 1700. In some aspects, apparatus 1700 includes a receiving component 1702 and a sending component 1704, which may communicate with each other (e.g., via one or more buses and / or one or more other components). As shown, apparatus 1700 may use receiving component 1702 and sending component 1704 to communicate with another apparatus 1706 (such as a UE, a base station, a network node, a network entity, a core network node, or another wireless communication device). As further shown, apparatus 1700 may include a communication manager 170. Communication manager 170 may include a determining component 1708, etc.

[0319] In some aspects, the apparatus 1700 may be configured to perform Figure 8 , Fig. 9A , Fig. 9B and Fig.10 Additionally or alternatively, the apparatus 1700 may be configured to perform one or more of the processes described herein (such as Fig.14 In some aspects, Fig.17 The device 1700 and / or one or more components shown may include a combination of Figure 2 Additionally or alternatively, Fig.17 One or more of the components shown may be combined with Figure 2 Additionally or alternatively, one or more components in the component set may be implemented at least in part as software stored in a memory. For example, a component (or a portion of a component) may be implemented as instructions or code stored in a non-transitory computer-readable medium and can be executed by a controller or processor to perform the function or operation of the component.

[0320] The receiving component 1702 may receive communications, such as reference signals, control information, data communications, or combinations thereof, from the device 1706. The receiving component 1702 may provide the received communications to one or more other components of the device 1700. In some aspects, the receiving component 1702 may perform signal processing (such as filtering, amplification, demodulation, analog-to-digital conversion, demultiplexing, deinterleaving, demapping, equalization, interference cancellation, or decoding, etc.) on the received communications and may provide the processed signals to one or more other components of the device 1700. In some aspects, the receiving component 1702 may include combining Figure 1 One or more antennas, modems, demodulators, MIMO detectors, receive processors, controllers / processors, memories, or combinations thereof, of the described core network nodes.

[0321] The transmitting component 1704 may transmit communications, such as reference signals, control information, data communications, or combinations thereof, to the device 1706. In some aspects, one or more other components of the device 1700 may generate communications and may provide the generated communications to the transmitting component 1704 for transmission to the device 1706. In some aspects, the transmitting component 1704 may perform signal processing (such as filtering, amplification, modulation, digital-to-analog conversion, multiplexing, interleaving, mapping, or encoding, etc.) on the generated communications and may transmit the processed signals to the device 1706. In some aspects, the transmitting component 1704 may include combining Figure 1 One or more antennas, modems, modulators, transmit MIMO processors, transmit processors, controllers / processors, memories, or combinations thereof of the described core network nodes. In some aspects, the transmit component 1704 can be co-located with the receive component 1702 in a transceiver.

[0322] Receiving component 1702 may receive a first handover request communication associated with a handover procedure for a UE from a first network node, wherein the first handover request communication indicates at least one of first machine learning inference information associated with a machine learning service, machine learning data associated with the UE, or a first inference host associated with the first network node. Sending component 1704 may send a second handover request communication associated with a handover procedure for the UE to a second network node, wherein the second network node is selected for the handover procedure based at least in part on at least one of the first machine learning inference information or a machine learning capability associated with the second network node. Receiving component 1702 may receive a first handover response communication associated with the handover procedure from the second network node, the first handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node.

[0323] Determining component 1708 can select the second network node for the handover process based at least in part on at least one of the first machine learning inference information, machine learning capabilities associated with the second network node, or capabilities associated with the second inference host.

[0324] The sending component 1704 can send a second handover response communication associated with the handover process to the first network node, the second handover response communication indicating second machine learning inference information.

[0325] Fig.17 The number and arrangement of components shown are provided as examples. In practice, there may be Fig.17 Additional components, fewer components, different components, or components arranged in a different manner than those shown. Fig.17 Two or more components shown may be implemented in a single component, or Fig.17 The single component shown may be implemented as multiple distributed components. Additionally or alternatively, Fig.17 The component set (one or more components) shown may be described as being executable by Fig.17 Another collection of components shown performs one or more functions.

[0326] The following provides an overview of some aspects of the disclosure:

[0327] Aspect 1: A method of wireless communication performed by a user equipment (UE), the method comprising: sending first machine learning data associated with a machine learning service to a first network node for use by a first inference host associated with the first network node; receiving a handover command communication from the first network node, the handover command communication indicating that the UE will perform a handover from the first network node to a second network node, wherein the handover command communication indicates machine learning inference information associated with a second inference host associated with the second network node; and sending second machine learning data associated with the machine learning service to the second network node for use by the second inference host associated with the second network node based at least in part on receiving the handover command communication.

[0328] Aspect 2: The method according to aspect 1, wherein the UE is associated with a vehicle.

[0329] Aspect 3: The method according to any one of Aspects 1 to 2 further includes: obtaining one or more sensor measurements associated with one or more objects via one or more sensors; and processing the one or more sensor measurements via a machine learning model executed on the UE to obtain the machine learning data associated with the machine learning service.

[0330] Aspect 4: A method according to any one of Aspects 1 to 3, wherein at least one of the first machine learning data or the second machine learning data includes at least one of the following items: the geographic location of the UE, the direction of travel of the UE, the speed of the UE, beam information associated with another UE co-located with the UE, or data extracted from one or more sensor measurements.

[0331] Aspect 5: The method according to any one of Aspects 1 to 4 further includes: sending a random access channel (RACH) connection communication or a radio resource control (RRC) connection communication to the second network node indicating at least one of the following items: information associated with the first machine learning data, information associated with one or more machine learning models used to generate the first machine learning data, information associated with the machine learning service, an indication of one or more machine learning service types associated with the machine learning service, one or more machine learning model identifiers associated with the machine learning service, or one or more machine learning inference host identifiers associated with the machine learning service.

[0332] Aspect 6: The method according to any one of Aspects 1 to 5 further includes: receiving a first protocol data unit (PDU) session communication from the second network node, the first protocol data unit (PDU) session communication indicating PDU information associated with the machine learning service; and in response to the first PDU session communication, sending a second PDU session communication to the second network node to establish a PDU session with the second inference host.

[0333] Aspect 7: A method according to any one of Aspects 1 to 6, wherein sending the second machine learning data includes: processing one or more sensor measurements via a machine learning model executed on the UE to obtain the second machine learning data associated with the machine learning service; and sending the second machine learning data to the second network node or the second inference host via a protocol data unit (PDU) session.

[0334] Aspect 8: A method according to any one of Aspects 1 to 7, wherein the UE is co-located with another UE, and wherein the machine learning data associated with the machine learning service is associated with the other UE.

[0335] Aspect 9: The method according to any one of Aspects 1 to 8, wherein the machine learning service comprises at least one of the following: beam prediction, handover prediction or beam blocking prediction.

[0336] Aspect 10: A method of wireless communication performed by a first network node, the method comprising: receiving a first handover communication from a first inference host associated with the first network node, the first handover communication indicating first machine learning inference information associated with a machine learning service, the machine learning service being associated with machine learning data associated with a first user equipment (UE); sending a handover request communication associated with a handover process for the first UE to a second network node or a core network node, wherein the first network node selects the second network node for the handover process based at least in part on the first machine learning inference information and the machine learning service associated with the second network node; and receiving a handover response communication associated with the handover process from the second network node or the core network node, the handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node.

[0337] Aspect 11: The method according to Aspect 10 also includes: receiving a measurement report from the first UE or the second UE, the measurement report indicating one or more measurements associated with one or more candidate network nodes including the second network node; and determining to hand over at least one of the first UE or the second UE to the second network node based at least in part on the measurement report, the first machine learning inference information, and the machine learning service associated with the second network node.

[0338] Aspect 12: The method according to any one of Aspects 10 to 11 further includes: sending a second handover communication indicating at least one of machine learning service information or machine learning data to the first inference host, wherein the first handover communication is in response to the second handover communication.

[0339] Aspect 13: A method according to Aspect 12, wherein the machine learning data includes at least one of the following: an identifier of a machine learning data service provided by the first UE, an identifier associated with the first UE, an indication of one or more machine learning models associated with the machine learning data service executed on the first UE, or an indication of one or more machine learning data types associated with one or more machine learning operations performed by the first UE for the machine learning data service.

[0340] Aspect 14: A method according to any one of Aspects 10 to 13, wherein the first machine learning inference information indicates at least one of the following: one or more machine learning models associated with the machine learning service executed on the first inference host, or information associated with the machine learning service or the state of the machine learning data associated with the machine learning service.

[0341] Aspect 15: The method according to any one of aspects 10 to 14, wherein the handover request communication indicates the first machine learning inference information received from the first inference host.

[0342] Aspect 16: The method according to any one of aspects 10 to 15, wherein the handover request communication is sent via an Xn interface.

[0343] Aspect 17: The method according to any one of Aspects 10 to 16 also includes: sending a third handover communication to the first inference host based at least in part on receiving the handover response communication, wherein the handover response communication indicates that the handover of at least one of the first UE or the second UE to the second network node is confirmed.

[0344] Aspect 18: The method according to any one of Aspects 10 to 17 further includes: sending a radio resource control (RRC) communication indicating a handover command for the first UE to the first UE, wherein the RRC communication indicates the second machine learning inference information associated with the second inference host.

[0345] Aspect 19: The method according to any one of Aspects 10 to 18 further includes: after sending the handover request communication, receiving machine learning data for the machine learning service associated with the first UE from the first inference host; sending the machine learning data for the machine learning service associated with the first UE to the second network node or the core network node; and sending context information associated with the first UE to the second network node or the core network node, the context information including at least one of machine learning service information or machine learning data.

[0346] Aspect 20: The method according to any one of Aspects 10 to 19 further includes: sending UE context information and data associated with a second UE co-located with the first UE to the second network node; receiving one or more machine learning predictions associated with the second UE from the second network node, wherein the one or more machine learning predictions are at least partially based on machine learning data generated by the first UE; and sending a radio resource control (RRC) communication indicating a handover command to the second UE, wherein the RRC communication indicates that the second UE will connect to the second network node based on the one or more machine learning predictions.

[0347] Aspect 21: A method according to any one of aspects 10 to 20, wherein the core network node is associated with an access and mobility function (AMF).

[0348] Aspect 22: A method of wireless communication performed by a first network node, the method comprising: receiving a handover request communication associated with a handover process for a user equipment (UE) from a second network node or a core network node, wherein the handover request communication indicates first machine learning inference information associated with a machine learning service and machine learning data associated with the UE; communicating with a first inference host associated with the first network node to determine that the first inference host is capable of supporting the machine learning service; and sending a handover response communication indicating second machine learning inference information associated with the first inference host to the second network node or the core network node based at least in part on the first inference host being able to support the machine learning service.

[0349] Aspect 23: A method according to Aspect 22, wherein communicating to determine whether the first inference host is capable of supporting the machine learning service includes: sending a first handover communication indicating at least one of the first machine learning inference information or the machine learning data to the first inference host; and receiving a second handover communication from the first inference host indicating whether the first inference host is capable of supporting the machine learning service, wherein the second handover communication indicates the second machine learning inference information associated with the first inference host.

[0350] Aspect 24: A method according to any one of Aspects 22 to 23, wherein the first machine learning inference information indicates at least one of the following: an identifier of a second inference host associated with the machine learning service and the second network node, a machine learning model executed on the second inference host, or other machine learning data associated with the machine learning service associated with the second inference host.

[0351] Aspect 25: A method according to any one of Aspects 22 to 24, wherein the second machine learning inference information indicates at least one of the following: an identifier of the first inference host, a machine learning model executed on the first inference host, or the machine learning data associated with the machine learning service associated with the first inference host.

[0352] Aspect 26: The method according to any one of Aspects 22 to 25 further includes: receiving a random access channel (RACH) connection communication or a radio resource control (RRC) connection communication indicating machine learning service information associated with the UE from the UE, the machine learning service information indicating at least one of the following: machine learning data associated with the machine learning service, machine learning type or service information associated with the machine learning service, or information associated with a machine learning model associated with the machine learning service executed on the UE.

[0353] Aspect 27: The method according to any one of Aspects 22 to 26 further includes: receiving UE context information from the second network node or the core network node, the UE context information including at least one of the machine learning service information or the machine learning data.

[0354] Aspect 28: The method according to any one of Aspects 22 to 27 also includes: sending a first protocol data unit (PDU) communication to the first inference host, the first protocol data unit (PDU) communication indicating the first machine learning inference information associated with the machine learning service for the UE; and receiving a second PDU communication from the first inference host, the second PDU communication indicating PDU session information of a PDU session associated with the machine learning service.

[0355] Aspect 29: The method according to Aspect 28 also includes: sending an indication of the PDU session information of the PDU session associated with the machine learning service to the UE.

[0356] Aspect 30: The method according to any one of Aspects 22 to 29 further includes: sending a protocol data unit (PDU) to the core network node to establish communication, wherein the protocol data unit (PDU) to establish communication indicates the first machine learning inference information associated with the machine learning service for the UE, so that the PDU session associated with the machine learning service is associated with the first inference host.

[0357] Aspect 31: The method according to any one of Aspects 22 to 30 further includes: after sending the handover response communication, receiving machine learning data associated with the UE for the machine learning service from the second network node; and sending the machine learning data to the first inference host.

[0358] Aspect 32: The method according to any one of Aspects 22 to 31 further includes: receiving one or more machine learning predictions associated with another UE from the first inference host, the one or more machine learning predictions being at least partially based on machine learning data provided by the UE; and sending an indication of the one or more machine learning predictions to the second network node so as to facilitate handing over the other UE to the first network node based on the one or more machine learning predictions.

[0359] Aspect 33: The method according to Aspect 32 also includes: receiving a radio resource control (RRC) connection request communication from the other UE based on the one or more machine learning predictions.

[0360] Aspect 34: A method according to any one of Aspects 32 to 33, wherein the one or more machine learning predictions include a beam or beam pair to be used by the other UE to establish a connection with the first network node.

[0361] Aspect 35: A method of wireless communication performed by a core network node, the method comprising: receiving a first handover request communication associated with a handover process for a user equipment (UE) from a first network node, wherein the first handover request communication indicates at least one of first machine learning inference information associated with a machine learning service, machine learning data associated with the UE, or a first inference host associated with the first network node; sending a second handover request communication associated with the handover process for the UE to a second network node, wherein the second network node is selected for the handover process based at least in part on the first machine learning inference information and a machine learning capability associated with the second network node; and receiving a first handover response communication associated with the handover process from the second network node, the first handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node.

[0362] Aspect 36: The method according to Aspect 35 also includes: selecting the second network node for the handover process based at least in part on at least one of the first machine learning inference information, the machine learning capabilities associated with the second network node, or the capabilities associated with the second inference host.

[0363] Aspect 37: The method according to any one of Aspects 35 to 36 further includes: sending a second handover response communication associated with the handover process to the first network node, the second handover response communication indicating the second machine learning inference information.

[0364] Aspect 38: A method according to any one of Aspects 35 to 37, wherein the first machine learning inference information includes at least one of the following: an identifier of the first inference host, an indication of a machine learning model associated with the machine learning service executed on the first inference host, or machine learning data associated with the machine learning service.

[0365] Aspect 39: A method according to any one of Aspects 35 to 38, wherein the machine learning data indicates at least one of: an identifier of a machine learning data service associated with the UE, a machine learning model associated with the machine learning service executed on the UE, or a machine learning data type or format associated with the machine learning service.

[0366] Aspect 40: A method according to any one of Aspects 35 to 39, wherein the second machine learning inference information indicates at least one of the following: an identifier of the second inference host, a machine learning model associated with the machine learning service executed on the second inference host, or machine learning data associated with the machine learning service.

[0367] Aspect 41: The method according to any one of aspects 35 to 40, wherein the first handover request communication is received via an N2 interface.

[0368] Aspect 42: A method according to any one of aspects 35 to 41, wherein the core network node is associated with an access and mobility function (AMF).

[0369] Aspect 43: An apparatus for wireless communication at a device, comprising: a processor; a memory coupled to the processor; and instructions, the instructions being stored in the memory and executable by the processor to cause the apparatus to perform one or more of the methods described in Aspects 1 to 9.

[0370] Aspect 44: A device for wireless communication, the device comprising: a memory; and one or more processors coupled to the memory, the one or more processors configured to perform the method according to one or more of aspects 1 to 9.

[0371] Aspect 45: An apparatus for wireless communication, the apparatus comprising: at least one means for performing the method according to one or more of aspects 1 to 9.

[0372] Aspect 46: A non-transitory computer-readable medium storing code for wireless communication, the code comprising instructions executable by a processor to perform the method according to one or more of aspects 1 to 9.

[0373] Aspect 47: A non-transitory computer-readable medium storing an instruction set for wireless communication, the instruction set comprising one or more instructions that, when executed by one or more processors of a device, cause the device to perform one or more of the methods described in aspects 1 to 9.

[0374] Aspect 48: An apparatus for performing wireless communications at a device, the apparatus comprising: a processor; a memory coupled to the processor; and instructions, the instructions being stored in the memory and executable by the processor to cause the apparatus to perform one or more of the methods described in Aspects 10 to 22 and / or Aspects 23 to 34.

[0375] Aspect 49: A device for wireless communication, the device comprising: a memory; and one or more processors coupled to the memory, the one or more processors being configured to execute the method according to one or more of Aspects 10 to 22 and / or Aspects 23 to 34.

[0376] Aspect 50: An apparatus for wireless communication, the apparatus comprising: at least one means for performing the method according to one or more of aspects 10 to 22 and / or aspects 23 to 34.

[0377] Aspect 51: A non-transitory computer-readable medium storing code for wireless communication, the code comprising instructions executable by a processor to perform the method according to one or more of aspects 10 to 22 and / or aspects 23 to 34.

[0378] Aspect 52: A non-transitory computer-readable medium storing an instruction set for wireless communication, the instruction set comprising one or more instructions that, when executed by one or more processors of a device, cause the device to perform one or more of the methods described in aspects 10 to 22 and / or aspects 23 to 34.

[0379] Aspect 53: An apparatus for wireless communication at a device, the apparatus comprising: a processor; a memory coupled to the processor; and instructions, the instructions being stored in the memory and executable by the processor to cause the apparatus to perform one or more of the methods described in Aspects 35 to 42.

[0380] Aspect 54: An apparatus for wireless communication, the apparatus comprising: a memory; and one or more processors coupled to the memory, the one or more processors configured to perform the method according to one or more of aspects 35 to 42.

[0381] Aspect 55: An apparatus for wireless communication, the apparatus comprising: at least one means for performing the method according to one or more of aspects 35 to 42.

[0382] Aspect 56: A non-transitory computer-readable medium storing code for wireless communication, the code comprising instructions executable by a processor to perform the method according to one or more of aspects 35 to 42.

[0383] Aspect 57: A non-transitory computer-readable medium storing an instruction set for wireless communication, the instruction set comprising one or more instructions that, when executed by one or more processors of a device, cause the device to perform one or more of the methods described in aspects 35 to 42.

[0384] The foregoing disclosure provides illustration and description, but is not intended to be exhaustive or to limit the aspects to the precise forms disclosed. Modifications and variations may be made in light of the above disclosure or may be acquired from practice of these aspects.

[0385] As used herein, the term "component" is intended to be broadly interpreted as a combination of hardware and / or hardware and software. Whether referred to as softwar...

Claims

1. A user equipment (UE) for wireless communication, the user equipment (UE) comprising: Memory; and one or more processors coupled to the memory and configured to: sending first machine learning data associated with the machine learning service to the first network node for use by a first inference host associated with the first network node; receiving a handover command communication from the first network node, the handover command communication indicating that the UE is to perform a handover from the first network node to a second network node, wherein the handover command communication indicates machine learning inference information associated with a second inference host associated with the second network node; as well as Based at least in part on receiving the handover command communication, second machine learning data associated with the machine learning service is sent to the second network node for use by the second inference host associated with the second network node.

2. The UE according to claim 1, wherein at least one of the first machine learning data or the second machine learning data comprises at least one of the following: the geographical location of the UE, the direction of travel of the UE, The speed of the UE, beam information associated with another UE co-located with the UE, or Data extracted from one or more sensor measurements.

3. The UE of claim 1 , wherein the one or more processors are further configured to: sending a random access channel (RACH) connection communication or a radio resource control (RRC) connection communication to the second network node indicating at least one of the following: information associated with the first machine learning data, information associated with one or more machine learning models used to generate the first machine learning data, information associated with the machine learning service, an indication of one or more machine learning service types associated with the machine learning service, one or more machine learning model identifiers associated with the machine learning service, or One or more machine learning inference host identifiers associated with the machine learning service.

4. The UE of claim 1, wherein the one or more processors are further configured to: receiving a first protocol data unit (PDU) session communication from the second network node, the first protocol data unit (PDU) session communication indicating PDU information associated with the machine learning service; and In response to the first PDU session communication, a second PDU session communication is sent to the second network node to establish a PDU session with the second inference host.

5. The UE of claim 1 , wherein to send the second machine learning data, the one or more processors are configured to: processing one or more sensor measurements via a machine learning model executed on the UE to obtain the second machine learning data associated with the machine learning service; and The second machine learning data is sent to the second network node or the second inference host via a protocol data unit (PDU) session.

6. The UE of claim 1, wherein the UE is co-located with another UE, and wherein the first machine learning data associated with the machine learning service is associated with the other UE.

7. A first network node for wireless communication, the network node comprising: Memory; and one or more processors coupled to the memory and configured to: receiving a first handover communication from a first inference host associated with the first network node, the first handover communication indicating first machine learning inference information associated with a machine learning service, the machine learning service associated with machine learning data associated with a first user equipment (UE); sending a handover request communication associated with a handover procedure for the first UE to a second network node or a core network node, wherein the first network node selects the second network node for the handover procedure based at least in part on the first machine learning inference information and a machine learning service associated with the second network node; as well as A handover response communication associated with the handover process is received from the second network node or the core network node, the handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node.

8. The first network node of claim 7, wherein the one or more processors are further configured to: receiving a measurement report from the first UE or the second UE, the measurement report indicating one or more measurements associated with one or more candidate network nodes including the second network node; and Determine to hand over at least one of the first UE or the second UE to the second network node based at least in part on the measurement report, the first machine learning inference information, and the machine learning service associated with the second network node.

9. The first network node of claim 7, wherein the one or more processors are further configured to: A second handover communication indicating at least one of machine learning service information or machine learning data is sent to the first inference host, wherein the first handover communication is responsive to the second handover communication.

10. The first network node of claim 9, wherein the machine learning data comprises at least one of the following: an identifier of a machine learning data service provided by the first UE, an identifier associated with the first UE, an indication of one or more machine learning models associated with the machine learning data service executed on the first UE, or An indication of one or more machine learning data types associated with one or more machine learning operations performed by the first UE for the machine learning data service.

11. The first network node of claim 7, wherein the first machine learning inference information indicates at least one of the following: one or more machine learning models associated with the machine learning service executed on the first inference host, or Information associated with the machine learning service or the state of the machine learning data associated with the machine learning service.

12. The first network node of claim 7, wherein the one or more processors are further configured to: Based at least in part on receiving the handover response communication, sending a third handover communication to the first inference host, the handover response communication indicating that handover of at least one of the first UE or the second UE to the second network node is confirmed.

13. The first network node of claim 7, wherein the one or more processors are further configured to: A radio resource control (RRC) communication is sent to the first UE indicating a handover command for the first UE, wherein the RRC communication indicates the second machine learning inference information associated with the second inference host.

14. The first network node of claim 7, wherein the one or more processors are further configured to: receiving, after sending the handover request communication, machine learning data for the machine learning service associated with the first UE from the first inference host; sending the machine learning data associated with the first UE for the machine learning service to the second network node or the core network node; and Send context information associated with the first UE to the second network node or the core network node, the context information comprising at least one of machine learning service information or machine learning data.

15. The first network node of claim 7, wherein the one or more processors are further configured to: sending, to the second network node, UE context information and data associated with a second UE co-located with the first UE; receiving, from the second network node, one or more machine learning predictions associated with the second UE, wherein the one or more machine learning predictions are based at least in part on machine learning data generated by the first UE; and A radio resource control (RRC) communication is sent to the second UE indicating a handover command, wherein the RRC communication indicates that the second UE is to connect with the second network node based on the one or more machine learning predictions.

16. A first network node for wireless communication, the first network node comprising: Memory; and one or more processors coupled to the memory and configured to: receiving a handover request communication associated with a handover procedure for a user equipment (UE) from a second network node or a core network node, wherein the handover request communication indicates first machine learning inference information associated with a machine learning service and machine learning data associated with the UE; communicating with a first inference host associated with the first network node to determine that the first inference host is capable of supporting the machine learning service; as well as Based at least in part on the first inference host being able to support the machine learning service, sending a handoff response communication indicating second machine learning inference information associated with the first inference host to the second network node or the core network node.

17. The first network node of claim 16, wherein to communicate to determine that the first inference host is capable of supporting the machine learning service, the one or more processors are configured to: sending a first handoff communication indicating at least one of the first machine learning inference information or the machine learning data to the first inference host; and A second handoff communication is received from the first inference host indicating whether the first inference host is capable of supporting the machine learning service, wherein the second handoff communication indicates the second machine learning inference information associated with the first inference host.

18. The first network node of claim 16, wherein the first machine learning inference information indicates at least one of the following: an identifier of a second inference host associated with the machine learning service and the second network node, executing a machine learning model on the second inference host, or Other machine learning data associated with the machine learning service associated with the second inference host.

19. The first network node of claim 16, wherein the one or more processors are further configured to: Receiving, from the UE, a random access channel (RACH) connection communication or a radio resource control (RRC) connection communication indicating machine learning service information associated with the UE, the machine learning service information indicating at least one of the following: machine learning data associated with the machine learning service, the machine learning type or service information associated with the machine learning service, or Information associated with a machine learning model associated with the machine learning service executed on the UE.

20. The first network node of claim 16, wherein the one or more processors are further configured to: UE context information is received from the second network node or the core network node, the UE context information comprising at least one of machine learning service information or the machine learning data.

21. The first network node of claim 16, wherein the one or more processors are further configured to: sending a first protocol data unit (PDU) communication to the first inference host, the first protocol data unit (PDU) communication indicating the first machine learning inference information associated with the machine learning service for the UE; and A second PDU communication is received from the first inference host, the second PDU communication indicating PDU session information for a PDU session associated with the machine learning service.

22. The first network node of claim 21, wherein the one or more processors are further configured to: Sending an indication of the PDU session information for the PDU session associated with the machine learning service to the UE.

23. The first network node of claim 16, wherein the one or more processors are further configured to: A protocol data unit (PDU) is sent to the core network node to establish communication, wherein the protocol data unit (PDU) establishes communication indicating the first machine learning inference information associated with the machine learning service for the UE so that a PDU session associated with the machine learning service is associated with the first inference host.

24. The first network node of claim 16, wherein the one or more processors are further configured to: receiving, after sending the handover response communication, machine learning data associated with the UE for the machine learning service from the second network node; and The machine learning data is sent to the first inference host.

25. The first network node of claim 16, wherein the one or more processors are further configured to: receiving, from the first inference host, one or more machine learning predictions associated with another UE, the one or more machine learning predictions being based at least in part on machine learning data provided by the UE; and and sending an indication of the one or more machine learning predictions to the second network node to facilitate handing over the other UE to the first network node based on the one or more machine learning predictions.

26. The first network node of claim 25, wherein the one or more processors are further configured to: A radio resource control (RRC) connection request communication is received from the other UE based on the one or more machine learning predictions.

27. The first network node of claim 25, wherein the one or more machine learning predictions comprise a beam or beam pair to be used by the other UE to establish a connection with the first network node.

28. A core network node for wireless communication, the core network node comprising: Memory; and one or more processors coupled to the memory and configured to: receiving, from a first network node, a first handover request communication associated with a handover procedure for a user equipment (UE), wherein the first handover request communication indicates at least one of first machine learning inference information associated with a machine learning service, machine learning data associated with the UE, or a first inference host associated with the first network node; sending a second handover request communication associated with the handover procedure for the UE to a second network node, wherein the second network node is selected for the handover process based at least in part on the first machine learning inference information and a machine learning capability associated with the second network node; as well as A first handover response communication associated with the handover process is received from the second network node, the first handover response communication indicating second machine learning inference information associated with a second inference host associated with the second network node.

29. The core network node of claim 28, wherein the one or more processors are further configured to: The second network node is selected for the handover process based at least in part on at least one of the first machine learning inference information, the machine learning capabilities associated with the second network node, or capabilities associated with the second inference host.

30. The core network node of claim 28, wherein the one or more processors are further configured to: A second handover response communication associated with the handover process is sent to the first network node, the second handover response communication indicating the second machine learning inference information.

Images