Node selection for radio frequency fingerprint (RFFP) federated learning

Abstract

Techniques for training machine learning models are disclosed. In one aspect, a user equipment (UE) receives, from a network entity, one or more selection criteria for determining whether the UE is to participate in training a machine learning model; determining whether the UE satisfies the one or more selection criteria during the first time period; and after the second time period, sending updated parameters of the machine learning model to the network entity, where the machine learning model is updated during the second time period based on a determination that the UE satisfies the one or more selection criteria.

Description

Technical Field

[0001] Aspects of the present disclosure generally relate to wireless communications. Background Art

[0002] Wireless communication systems have evolved over many generations, including first generation analog wireless telephone service (1G), second generation (2G) digital wireless telephone service (including transitional 2.5G and 2.75G networks), third generation (3G) high speed data, Internet-enabled wireless services, and fourth generation (4G) services (e.g., Long Term Evolution (LTE) or WiMax), etc. There are many different types of wireless communication systems in use today, including cellular and personal communications service (PCS) systems. Examples of known cellular systems include cellular analog Advanced Mobile Phone System (AMPS) and digital cellular systems based on code division multiple access (CDMA), frequency division multiple access (FDMA), time division multiple access (TDMA), Global System for Mobile Communications (GSM), etc.

[0003] The fifth generation (5G) wireless standard, known as New Radio (NR), enables higher data speeds, more connections, and better coverage, among other improvements. According to the Next Generation Mobile Networks Alliance, the 5G standard is designed to provide higher data rates than previous standards, more accurate positioning (e.g., based on positioning reference signals (RS-P), such as downlink, uplink, or sidelink positioning reference signals (PRS)), and other technology enhancements. These enhancements, along with the use of higher frequency bands, advances in PRS procedures and technologies, and high-density deployments of 5G, enable highly accurate 5G-based positioning. Summary of the invention

[0004] A brief summary of one or more aspects disclosed herein is presented below. Therefore, the following summary should not be considered an extensive review of all contemplated aspects, nor should it be considered to identify key or important elements related to all contemplated aspects or to delineate the scope associated with any particular aspect. Therefore, the sole purpose of the following summary is to present certain concepts related to one or more aspects of the mechanisms disclosed herein in a simplified form prior to the detailed description presented below.

[0005] In one aspect, a method for training a machine learning model performed by a user equipment (UE) includes: receiving one or more selection criteria from a network entity for determining whether the UE is to participate in training the machine learning model; determining whether the UE satisfies the one or more selection criteria during a first time period; and after a second time period, sending updated parameters of the machine learning model to the network entity, wherein the machine learning model is updated during the second time period based on a determination that the UE satisfies the one or more selection criteria.

[0006] In one aspect, a method for training a machine learning model performed by a network entity includes: sending one or more selection criteria to a set of user equipment (UEs) for determining whether the set of UEs is to participate in training the machine learning model; sending the machine learning model to at least one subset of UEs in the set of UEs that meet the one or more selection criteria; receiving updated parameters of the machine learning model from each UE in the subset of UEs; and updating the machine learning model based on the updated parameters received from each UE in the subset of UEs.

[0007] In one aspect, a user equipment (UE) includes: a memory; at least one transceiver; and at least one processor communicatively coupled to the memory and the at least one transceiver, the at least one processor being configured to: receive one or more selection criteria for determining whether the UE is to participate in training a machine learning model from a network entity via the at least one transceiver; determine whether the UE satisfies the one or more selection criteria during a first time period; and after a second time period, send updated parameters of the machine learning model to the network entity via the at least one transceiver, wherein the machine learning model is updated during the second time period based on a determination that the UE satisfies the one or more selection criteria.

[0008] In one aspect, a network entity includes: a memory; at least one transceiver; and at least one processor communicatively coupled to the memory and the at least one transceiver, the at least one processor being configured to: send one or more selection criteria for determining whether the set of UEs is to participate in training a machine learning model to a set of user equipment (UEs) via at least one transceiver; send the machine learning model to at least one subset of UEs in the set of UEs that meet the one or more selection criteria via at least one transceiver; receive updated parameters of the machine learning model from each UE in the subset of UEs via at least one transceiver; and update the machine learning model based on the updated parameters received from each UE in the subset of UEs.

[0009] In one aspect, a user equipment (UE) includes: a component for receiving one or more selection criteria from a network entity for determining whether the UE is to participate in training a machine learning model; a component for determining whether the UE satisfies the one or more selection criteria during a first time period; and a component for sending updated parameters of the machine learning model to the network entity after a second time period, wherein the machine learning model is updated during the second time period based on a determination that the UE satisfies the one or more selection criteria.

[0010] In one aspect, a network entity includes: a component for sending one or more selection criteria to a set of user equipment (UE) for determining whether the set of UEs is to participate in training a machine learning model; a component for sending the machine learning model to at least one subset of UEs in the set of UEs that meet the one or more selection criteria; a component for receiving updated parameters of the machine learning model from each UE in the subset of UEs; and a component for updating the machine learning model based on the updated parameters received from each UE in the subset of UEs.

[0011] In one aspect, a non-transitory computer-readable medium stores computer-executable instructions that, when executed by a user equipment (UE), cause the UE to: receive one or more selection criteria from a network entity for determining whether the UE is to participate in training a machine learning model; determine whether the UE satisfies the one or more selection criteria during a first time period; and after a second time period, send updated parameters of the machine learning model to the network entity, wherein the machine learning model is updated during the second time period based on a determination that the UE satisfies the one or more selection criteria.

[0012] In one aspect, a non-transitory computer-readable medium stores computer-executable instructions that, when executed by a network entity, cause the network entity to: send one or more selection criteria to a set of user equipment (UEs) for determining whether the set of UEs is to participate in training a machine learning model; send the machine learning model to at least one subset of UEs in the set of UEs that meet the one or more selection criteria; receive updated parameters of the machine learning model from each UE in the subset of UEs; and update the machine learning model based on the updated parameters received from each UE in the subset of UEs.

[0013] Other objects and advantages associated with the aspects disclosed herein will be apparent to those skilled in the art based on the accompanying drawings and detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The drawings are presented to aid in describing various aspects of the disclosure and are provided solely for the purpose of illustrating these aspects and not limiting the disclosure.

[0015] Figure 1 An example wireless communication system in accordance with aspects of the present disclosure is shown.

[0016] Figure 2A , Figure 2B and Figure 2C Example wireless network structures in accordance with aspects of the present disclosure are shown.

[0017] Figure 3A , Figure 3B and Figure 3Cis a simplified block diagram of several sample aspects of components that may be used in a user equipment (UE), a base station, and a network entity, respectively, and configured to support communications as taught herein.

[0018] Figure 4 Examples of various positioning methods supported in New Radio (NR) according to aspects of the present disclosure are shown.

[0019] Figure 5 is a diagram illustrating an example frame structure in accordance with aspects of the present disclosure.

[0020] Figure 6 is a diagram representing radio frequency (RF) channel estimation in accordance with aspects of the present disclosure.

[0021] Figure 7 An example neural network in accordance with aspects of the present disclosure is shown.

[0022] Figure 8 is a diagram illustrating the use of a machine learning model for RF fingerprint (RFFP) based positioning in accordance with aspects of the present disclosure.

[0023] Fig. 9 is a diagram illustrating an inference cycle for UE-based downlink RFFP (DL-RFFP) positioning according to aspects of the present disclosure.

[0024] Fig.10 An example call flow for UE-based DL-RFFP positioning according to aspects of the present disclosure is shown.

[0025] Fig.11 and Fig.12 Example methods of training machine learning models according to aspects of the present disclosure are shown. DETAILED DESCRIPTION

[0026] Some aspects of the present disclosure are provided in the following description and related drawings, which are directed to various examples provided for illustrative purposes. Alternative aspects may be designed without departing from the scope of the present disclosure. Additionally, in order not to obscure the relevant details of the present disclosure, well-known elements of the present disclosure will not be described in detail or omitted.

[0027] The words "exemplary" and / or "example" are used herein to mean "serving as an example, instance, or illustration." Any aspect described herein as "exemplary" and / or "example" is not necessarily to be construed as preferred or advantageous over other aspects. Likewise, the term "aspects of the disclosure" does not require that all aspects of the disclosure include the discussed feature, advantage, or mode of operation.

[0028] Those skilled in the art will appreciate that the information and signals described below may be represented using any of a variety of different techniques and technologies. For example, depending in part on the specific application, in part on the desired design, in part on the corresponding technology, etc., the data, instructions, commands, information, signals, bits, symbols, and chips referenced throughout the following description may be represented by voltages, currents, electromagnetic waves, magnetic fields or particles, optical fields or particles, or any combination thereof.

[0029] Further, many aspects are described in terms of a sequence of actions to be performed by, for example, an element of a computing device. It will be appreciated that the various actions described herein may be performed by a specific circuit (e.g., an application specific integrated circuit (ASIC)), by program instructions executed by one or more processors, or by a combination of the two. Additionally, the sequence of actions described herein may be considered to be implemented entirely in any form of non-transitory computer-readable storage medium, in which a corresponding set of computer instructions is stored, which, when executed, will cause or instruct the associated processor of the device to perform the functions described herein. Therefore, various aspects of the present disclosure may be implemented in a variety of different forms, all of which are considered to be within the scope of the claimed subject matter. In addition, for each aspect described herein, the corresponding form of any such aspect may be described herein as, for example, "logic" that is "configured to" perform the described actions.

[0030] As used herein, unless otherwise specified, the terms "user equipment (UE)" and "base station" are not intended to be specific or otherwise limited to any specific radio access technology (RAT). In general, a UE can be any wireless communication device (e.g., a mobile phone, a router, a tablet computer, a laptop computer, a consumer asset location device, a wearable device (e.g., a smart watch, glasses, augmented reality (AR) / virtual reality (VR) headsets, etc.), a vehicle (e.g., a car, a motorcycle, a bicycle, etc.), an Internet of Things device (IoT), etc.) that a user uses to communicate over a wireless communication network. A UE can be mobile or can be fixed (e.g., at certain times) and can communicate with a radio access network (RAN). As used herein, the term "UE" can be interchangeably referred to as an "access terminal" or "AT", "client device", "wireless device", "subscriber equipment", "subscriber terminal", "subscriber station", "user terminal" or "UT", "mobile device", "mobile terminal", "mobile station" or variants thereof. Typically, a UE can communicate with a core network via a RAN, and through the core network, the UE can connect to external networks (such as the Internet) and other UEs. Of course, other mechanisms for the UE to connect to the core network and / or the Internet are also possible, such as through a wired access network, a wireless local area network (WLAN) network (e.g., based on the Institute of Electrical and Electronics Engineers (IEEE) 802.11 specification, etc.), etc.

[0031] A base station may operate according to one of several RATs for communicating with a UE, depending on the network in which it is deployed, and may alternatively be referred to as an access point (AP), a network node, a NodeB, an evolved NodeB (eNB), a next generation eNB (ng-eNB), a new radio (NR) NodeB (may also be referred to as a gNB or gNodeB), etc. A base station may be primarily used to support wireless access for a UE, including supporting data, voice, and / or signaling connections for the supported UE. In some systems, a base station may provide a pure edge node signaling function, while in other systems, additional control and / or network management functions may be provided. A communication link through which a UE may transmit a signal to a base station is referred to as an uplink (UL) channel (e.g., a reverse traffic channel, a reverse control channel, an access channel, etc.). A communication link through which a base station may transmit a signal to a UE is referred to as a downlink (DL) or forward link channel (e.g., a paging channel, a control channel, a broadcast channel, a forward traffic channel, etc.). The term traffic channel (TCH) as used herein may refer to an uplink / reverse or downlink / forward traffic channel.

[0032] The term "base station" may refer to a single physical transmit / receive point (TRP) or multiple physical TRPs, which may or may not be co-located. For example, where the term "base station" refers to a single physical TRP, the physical TRP may be the antenna of the base station corresponding to the cell (or several cell sectors) in which the base station is located. Where the term "base station" refers to multiple co-located physical TRPs, the physical TRP may be the antenna array of the base station (for example, in a multiple-input multiple-output (MIMO) system or where the base station employs beamforming). Where the term "base station" refers to multiple non-co-located physical TRPs, the physical TRP may be a distributed antenna system (DAS) (a spatially separated antenna network connected to a common source via a transmission medium) or a remote radio head (RRH) (a remote base station connected to a serving base station). Alternatively, the non-co-located physical TRP may be a serving base station that receives measurement reports from a UE and a neighboring base station whose reference radio frequency (RF) signal the UE is measuring. As used herein, because a TRP is the point at which a base station transmits and receives wireless signals, references to transmissions from a base station or receptions at a base station should be understood to refer to a specific TRP of a base station.

[0033] In some embodiments of supporting positioning of a UE, a base station may not support wireless access of the UE (e.g., may not support data, voice, and / or signaling connections of the UE), but instead may send a reference signal to the UE for measurement by the UE, and / or may receive and measure a signal sent by the UE. Such a base station may be referred to as a positioning beacon (e.g., when sending a signal to the UE) and / or a position measurement unit (e.g., when receiving and measuring a signal from the UE).

[0034] An "RF signal" includes an electromagnetic wave of a given frequency that transmits information through the space between a transmitter and a receiver. As used herein, a transmitter may transmit a single "RF signal" or multiple "RF signals" to a receiver. However, due to the propagation characteristics of RF signals through multipath channels, a receiver may receive multiple "RF signals" corresponding to each transmitted RF signal. The same transmitted RF signal on different paths between a transmitter and a receiver may be referred to as a "multipath" RF signal. As used herein, an RF signal may also be referred to as a "wireless signal" or simply a "signal," where, as is apparent from the context, the term "signal" refers to either a wireless signal or an RF signal.

[0035] Figure 1An example wireless communication system 100 according to various aspects of the present disclosure is shown. The wireless communication system 100 (which may also be referred to as a wireless wide area network (WWAN)) may include various base stations 102 (labeled as BS) and various UEs 104. The base stations 102 may include macro cell base stations (high power cellular base stations) and / or small cell base stations (low power cellular base stations). In one aspect, the macro cell base station may include an eNB and / or ng-eNB corresponding to an LTE network of the wireless communication system 100, or a gNB corresponding to an NR network of the wireless communication system 100, or a combination of the two, and the small cell base station may include a femto cell, a pico cell, a micro cell, etc.

[0036] The base stations 102 may collectively form a RAN and interface with a core network 170 (e.g., an evolved packet core (EPC) or a 5G core (5GC)) through a backhaul link 122, and interface with one or more location servers 172 (e.g., a location management function (LMF) or a secure user plane location (SUPL) location platform (SLP)) through the core network 170. The location server 172 may be part of the core network 170, or may be external to the core network 170. The location server 172 may be integrated with the base station 102. The UE 104 may communicate with the location server 172 directly or indirectly. For example, the UE 104 may communicate with the location server 172 via the base station 102 currently serving the UE 104. The UE 104 may also communicate with the location server 172 through another path, such as via an application server (not shown), via another network (such as via a wireless local area network (WLAN) access point (AP) (e.g., AP 150 described below)), and the like. For signaling purposes, communication between UE 104 and location server 172 may be represented as an indirect connection (e.g., through core network 170, etc.) or a direct connection (e.g., via direct connection 128 as shown), with intermediate nodes (if any) omitted from the signaling diagram for clarity.

[0037] The base stations 102 may perform, among other functions, functions related to one or more of transmitting user data, radio channel encryption and decryption, integrity protection, header compression, mobility control functions (e.g., handover, dual connectivity), inter-cell interference coordination, connection establishment and release, load balancing, distribution of non-access stratum (NAS) messages, NAS node selection, synchronization, RAN sharing, multimedia broadcast multicast service (MBMS), user and device tracking, RAN information management (RIM), paging, positioning, and warning messaging. The base stations 102 may communicate with each other directly or indirectly (e.g., via EPC / 5GC) via a backhaul link 134, which may be wired or wireless.

[0038] Base station 102 can communicate wirelessly with UE 104. Each base station 102 can provide communication coverage for a corresponding geographic coverage area 110. In one aspect, the base station 102 in each geographic coverage area 110 can support one or more cells. A "cell" is a logical communication entity for communicating with a base station (e.g., through some frequency resources, referred to as carrier frequency, component carrier, carrier, frequency band, etc.), and can be associated with an identifier (e.g., a physical cell identifier (PCI), an enhanced cell identifier (ECI), a virtual cell identifier (VCI), a cell global identifier (CGI), etc.) to distinguish cells operating via the same or different carrier frequencies. In some cases, different cells can be configured according to different protocol types (e.g., machine type communication (MTC), narrowband IoT (NB-IoT), enhanced mobile broadband (eMBB) or other) that can provide access to different types of UEs. Because a cell is supported by a specific base station, the term "cell" can refer to one or both of the logical communication entity and the base station that supports it, depending on the context. In addition, because a TRP is typically a physical transmission point of a cell, the terms "cell" and "TRP" can be used interchangeably. In some cases, the term “cell” may also refer to a geographic coverage area (eg, sector) of a base station, as long as a carrier frequency can be detected and used for communications within some portion of the geographic coverage area 110 .

[0039] Although the geographic coverage areas 110 of neighboring macrocell base stations 102 may partially overlap (e.g., in a handoff region), some geographic coverage areas 110 may be substantially overlapped by a larger geographic coverage area 110. For example, a small cell base station 102' (labeled "SC" for "small cell") may have a coverage area 110' that substantially overlaps with the geographic coverage areas 110 of one or more macrocell base stations 102. A network including small cell base stations and macrocell base stations may be referred to as a heterogeneous network. A heterogeneous network may also include a home eNB (HeNB), which may provide services to a restricted group referred to as a closed subscriber group (CSG).

[0040] The communication link 120 between the base station 102 and the UE 104 may include an uplink (also referred to as a reverse link) transmission from the UE 104 to the base station 102 and / or a downlink (DL) (also referred to as a forward link) transmission from the base station 102 to the UE 104. The communication link 120 may use MIMO antenna technology, including spatial multiplexing, beamforming, and / or transmit diversity. The communication link 120 may be over one or more carrier frequencies. The allocation of carriers may be asymmetric with respect to the downlink and uplink (e.g., the downlink may be allocated more or fewer carriers than the uplink).

[0041] The wireless communication system 100 may also include a wireless local area network (WLAN) access point (AP) 150 that communicates with a WLAN station (STA) 152 in an unlicensed spectrum (e.g., 5 GHz) via a communication link 154. When communicating in the unlicensed spectrum, the WLAN STA 152 and / or the WLAN AP 150 may perform a clear channel assessment (CCA) or a listen before talk procedure prior to communication to determine whether a channel is available.

[0042] The small cell base station 102' can operate in a licensed and / or unlicensed spectrum. When operating in an unlicensed spectrum, the small cell base station 102' can adopt LTE or NR technology and use the same 5GHz unlicensed spectrum used by the WLAN AP 150. The small cell base station 102' adopting LTE / 5G in the unlicensed spectrum can improve the coverage of the access network and / or increase the capacity of the access network. NR in the unlicensed spectrum can be referred to as NR-U. LTE in the unlicensed spectrum can be referred to as LTE-U, License Assisted Access (LAA) or MulteFire.

[0043] The wireless communication system 100 may also include a millimeter wave (mmW) base station 180 that can operate at mmW frequencies and / or near mmW frequencies to communicate with UE 182. Extremely high frequency (EHF) is part of the RF in the electromagnetic spectrum. The frequency range of EHF is 30GHz to 300GHz, and the wavelength is between 1 mm and 10 mm. The radio waves in this frequency band can be called millimeter waves. Near mmW can extend to a frequency of 3GHz with a wavelength of 100 mm. The super high frequency (SHF) band extends between 3GHz and 30GHz and is also called centimeter waves. Communications using mmW / near mmW radio frequency bands have high path loss and relatively short distances. The mmW base station 180 and UE 182 can use beamforming (transmitting and / or receiving) on ​​the mmW communication link 184 to compensate for the extremely high path loss and short distance. Further, it will be understood that in an alternative configuration, one or more base stations 102 can also use mmW or near mmW and beamforming to transmit. Therefore, it will be understood that the above description is merely exemplary, and should not be construed as limiting the various aspects disclosed herein.

[0044] Transmit beamforming is a technique for focusing an RF signal in a specific direction. Traditionally, when a network node (e.g., a base station) broadcasts an RF signal, it broadcasts the signal in all directions (omnidirectional). With transmit beamforming, the network node determines the location of a given target device (e.g., a UE) (relative to the transmitting network node) and projects a stronger downlink RF signal in that specific direction, thereby providing a faster (in terms of data rate) and stronger RF signal to the receiving device. In order to change the directionality of the RF signal when transmitting, the network node can control the phase and relative amplitude of the RF signal at each of the one or more transmitters that broadcast the RF signal. For example, the network node can use an antenna array (referred to as a "phased array" or "antenna array") that produces an RF beam that can be "steered" to point in different directions without actually moving the antenna. Specifically, the RF current from the transmitter is fed to separate antennas in the correct phase relationship so that the radio waves from different antennas are added together to increase the radiation in the desired direction while canceling to suppress the radiation in the undesired direction.

[0045] The transmit beams can be quasi-co-located, which means that they have the same parameters for the receiver (e.g., UE) regardless of whether the transmit antennas of the network nodes themselves are physically co-located. In NR, there are four types of quasi-co-located (QCL) relationships. Specifically, a given type of QCL relationship means that certain parameters about the second reference RF signal on the second beam can be derived from information about the source reference RF signal on the source beam. Therefore, if the source reference RF signal is QCL type A, the receiver can use the source reference RF signal to estimate the Doppler shift, Doppler spread, average delay, and delay spread of the second reference RF signal sent on the same channel. If the source reference RF signal is QCL type B, the receiver can use the source reference RF signal to estimate the Doppler shift and Doppler spread of the second reference RF signal sent on the same channel. If the source reference RF signal is QCL type C, the receiver can use the source reference RF signal to estimate the Doppler shift and average delay of the second reference RF signal sent on the same channel. If the source reference RF signal is QCL type D, the receiver can use the source reference RF signal to estimate the spatial reception parameters of the second reference RF signal sent on the same channel.

[0046] In receive beamforming, a receiver uses a receive beam to amplify an RF signal detected on a given channel. For example, a receiver may increase a gain setting and / or adjust a phase setting of an antenna array in a particular direction to amplify (e.g., increase its gain level) RF signals received from that direction. Thus, when a receiver is said to be beamforming in a certain direction, this means that the beam gain in that direction is high relative to the beam gain along other directions, or that the beam gain in that direction is the highest compared to the beam gain in that direction of all other receive beams available to the receiver. This results in a stronger received signal strength (e.g., reference signal received power (RSRP), reference signal received quality (RSRQ), signal-to-interference-plus-noise ratio (SINR), etc.) for the RF signal received from that direction.

[0047] The transmit beam and the receive beam may be spatially correlated. The spatial relationship means that parameters of a second beam (e.g., a transmit beam or a receive beam) of a second reference signal may be derived from information about a first beam (e.g., a receive beam or a transmit beam) of a first reference signal. For example, a UE may receive a reference downlink reference signal (e.g., a synchronization signal block (SSB)) from a base station using a specific receive beam. The UE may then form a transmit beam for sending an uplink reference signal (e.g., a sounding reference signal (SRS)) to the base station based on the receive beam parameters.

[0048] It should be noted that a "downlink" beam can be a transmit beam or a receive beam, depending on the entity forming the beam. For example, if the base station is forming a downlink beam to send a reference signal to the UE, the downlink beam is a transmit beam. However, if the UE is forming a downlink beam, the beam is a receive beam that receives the downlink reference signal. Similarly, an "uplink" beam can be a transmit beam or a receive beam, depending on the entity forming the beam. For example, if the base station is forming an uplink beam, the beam is an uplink receive beam, and if the UE is forming an uplink beam, the beam is an uplink transmit beam.

[0049] The electromagnetic spectrum is often subdivided into various categories, bands, channels, etc. based on frequency / wavelength. In 5G NR, the two initial operating bands are identified with the frequency range names FR1 (410MHz–7.125GHz) and FR2 (24.25GHz–52.6GHz). It should be understood that although a portion of FR1 is greater than 6GHz, FR1 is often (interchangeably) referred to as the "Sub-6 GHz" band in various documents and articles. A similar naming issue sometimes arises with respect to FR2, which is often (interchangeably) referred to as the millimeter wave band in documents and articles, although FR2 is different from the extremely high frequency (EHF) band (30GHz–300GHz) identified as the "millimeter wave" band by the International Telecommunication Union (ITU).

[0050] Frequencies between FR1 and FR2 are often referred to as mid-band frequencies. Recent 5G NR research has identified operating bands for these mid-band frequencies as the frequency range designation FR3 (7.125 GHz–24.25 GHz). Frequency bands that fall within FR3 may inherit FR1 characteristics and / or FR2 characteristics, and thus may effectively extend the characteristics of FR1 and / or FR2 to mid-band frequencies. In addition, higher frequency bands are currently being explored to extend 5G NR operations above 52.6 GHz. For example, three higher operating frequency bands have been identified as the frequency range designations FR4a or FR4-1 (52.6 GHz–71 GHz), FR4 (52.6 GHz–114.25 GHz), and FR5 (114.25 GHz–300 GHz). Each of these higher frequency bands falls within the EHF band.

[0051] In view of the above, unless otherwise stated, it should be understood that the term "sub-6GHz" and the like, if used herein, can broadly refer to frequencies below 6GHz, can be within FR1, or can include mid-band frequencies. Further, unless otherwise stated, it should be understood that the term "millimeter wave" and the like, if used herein, can broadly refer to frequencies that can include mid-band frequencies, can be within FR2, FR4, FR4-a or FR4-1 and / or FR5, or can be within the EHF band.

[0052] In a multi-carrier system, such as 5G, one of the carrier frequencies is referred to as the "primary carrier" or "anchor carrier" or "primary serving cell" or "PCell", while the remaining carrier frequencies are referred to as "secondary carriers" or "secondary serving cells" or "SCells". In carrier aggregation, the anchor carrier is a carrier operating on the primary frequency (e.g., FR1) used by the UE 104 / 182 and the cell in which the UE 104 / 182 performs an initial radio resource control (RRC) connection establishment procedure or initiates an RRC connection reestablishment procedure. The primary carrier carries all common and UE-specific control channels and can be a carrier in a licensed frequency (however, this is not always the case). A secondary carrier is a carrier operating on a second frequency (e.g., FR2), which can be configured once an RRC connection is established between the UE 104 and the anchor carrier, and which can be used to provide additional radio resources. In some cases, the secondary carrier can be a carrier in an unlicensed frequency. The secondary carrier may contain only necessary signaling information and signals, for example, those UE-specific information and signals may not be present in the secondary carrier, because the primary uplink and downlink carriers are typically UE-specific. This means that different UEs 104 / 182 in a cell can have different downlink primary carriers. The same is true for the uplink primary carrier. The network is able to change the primary carrier of any UE 104 / 182 at any time. For example, this is done to balance the load on different carriers. Because a "serving cell" (whether PCell or SCell) corresponds to a carrier frequency / component carrier with which a base station is communicating, the terms "cell", "serving cell", "component carrier", "carrier frequency", etc. can be used interchangeably.

[0053] For example, still referring to Figure 1 , one of the frequencies used by the macrocell base station 102 may be an anchor carrier (or "PCell"), while the other frequencies used by the macrocell base station 102 and / or the mmW base station 180 may be secondary carriers ("SCells"). Simultaneous transmission and / or reception of multiple carriers enables the UE 104 / 182 to significantly increase its data transmission and / or reception rate. For example, two 20 MHz aggregated carriers in a multi-carrier system would theoretically result in a two-fold increase in data rate (i.e., 40 MHz) compared to a single 20 MHz carrier.

[0054] The wireless communication system 100 may also include a UE 164, which may communicate with the macrocell base station 102 via a communication link 120 and / or communicate with the mmW base station 180 via a mmW communication link 184. For example, the macrocell base station 102 may support one PCell and one or more SCells for the UE 164, and the mmW base station 180 may support one or more SCells for the UE 164.

[0055] In some cases, UE 164 and UE 182 are capable of sidelink communication. A sidelink-enabled UE (SL-UE) can communicate with base station 102 via communication link 120 using a Uu interface (i.e., an air interface between a UE and a base station). SL-UEs (e.g., UE 164, UE 182) can also communicate directly with each other via a wireless sidelink 160 using a PC5 interface (i.e., an air interface between sidelink-enabled UEs). A wireless sidelink (or simply "sidelink") is an improvement to a core cellular (e.g., LTE, NR) standard that allows direct communication between two or more UEs without going through a base station. Sidelink communications can be unicast or multicast and can be used for device-to-device (D2D) media sharing, vehicle-to-vehicle (V2V) communications, vehicle-to-vehicle (V2X) communications (e.g., cellular V2X (cV2X) communications, enhanced V2X (eV2X) communications, etc.), emergency rescue applications, etc. One or more of a group of SL-UEs utilizing sidelink communications may be within the geographic coverage area 110 of the base station 102. Other SL-UEs in such a group may be outside the geographic coverage area 110 of the base station 102 or unable to receive transmissions from the base station 102. In some cases, a group of SL-UEs communicating via sidelink communications may utilize a one-to-many (1:M) system, in which each SL-UE transmits to each other SL-UE in the group. In some cases, the base station 102 facilitates resource scheduling for the sidelink communications. In other cases, the sidelink communications are performed between the SL-UEs without involving the base station 102.

[0056] In one aspect, the sidelink 160 can operate over a wireless communication medium of interest that can be shared with other vehicles and / or infrastructure access points and other wireless communications between other RATs. A "medium" can consist of one or more time, frequency, and / or spatial communication resources (e.g., comprising one or more channels across one or more carriers) associated with wireless communications between one or more transmitter / receiver pairs. In one aspect, the medium of interest can correspond to at least a portion of an unlicensed band shared between various RATs. Although different licensed bands have been reserved for certain communication systems (e.g., by government entities such as the Federal Communications Commission (FCC) in the United States), these systems, particularly those that employ small cell access points, have recently expanded operations to unlicensed bands such as the unlicensed national information infrastructure (U-NII) band used by wireless local area network (WLAN) technologies, most notably the IEEE 802.11x WLAN technology commonly referred to as "Wi-Fi". Example systems of this type include different variants of CDMA systems, TDMA systems, FDMA systems, orthogonal FDMA (OFDMA) systems, single carrier FDMA (SC-FDMA) systems, and the like.

[0057] It should be noted that although Figure 1 Only two of the UEs are shown as SL-UEs (i.e., UE 164 and UE 182), but any of the shown UEs may be SL-UEs. Further, although only UE 182 is described as being capable of beamforming, any of the shown UEs, including UE 164, may be capable of beamforming. Where SL-UEs are capable of beamforming, they may beamform toward each other (i.e., toward other SL-UEs), toward other UEs (e.g., UE 104), toward a base station (e.g., base station 102, 180, small cell 102', access point 150), and the like. Thus, in some cases, UE 164 and UE 182 may utilize beamforming on sidelink 160.

[0058] exist Figure 1 In the example of Figure 1104) can receive a signal 124 from one or more earth orbiting spacecraft (SV) 112 (e.g., satellites). In one aspect, SV 112 can be part of a satellite positioning system that UE 104 can use as an independent source of location information. A satellite positioning system typically includes a transmitter system (e.g., SV 112) positioned so that a receiver (e.g., UE 104) can determine their location on or above the earth based at least in part on a positioning signal (e.g., signal 124) received from the transmitter. Such a transmitter typically transmits a signal marked with a repeating pseudo-random noise (PN) code of a set number of chips. Although typically located in SV 112, the transmitter can sometimes be located on a ground-based control station, base station 102, and / or other UE 104. UE 104 can include one or more dedicated receivers that are specifically designed to receive signal 124 for deriving geographic location information from SV 112.

[0059] In a satellite positioning system, the use of signal 124 may be enhanced by various satellite-based augmentation systems (SBAS), which may be associated with or otherwise capable of use with one or more global and / or regional navigation satellite systems. For example, SBAS may include augmentation systems that provide integrity information, differential corrections, etc., such as Wide Area Augmentation System (WAAS), European Geostationary Navigation Overlay Service (EGNOS), Multifunction Satellite Augmentation System (MSAS), Global Positioning System (GPS) Assisted Geographic Augmentation Navigation or GPS and Geographic Augmentation Navigation System (GAGAN), etc. Thus, as used herein, a satellite positioning system may include any combination of one or more global and / or regional navigation satellites associated with such one or more satellite positioning systems.

[0060] In one aspect, additionally or alternatively, SV 112 can be part of one or more non-terrestrial networks (NTNs). In an NTN, SV 112 is connected to an earth station (also referred to as a ground station, NTN gateway, or gateway), which in turn is connected to an element in a 5G network, such as an improved base station 102 (without a ground antenna) or a network node in a 5GC. This element in turn will provide access to other elements in the 5G network, and ultimately provide access to entities outside the 5G network, such as Internet web servers and other user devices. In this way, instead of or in addition to receiving communication signals from a ground base station 102, UE 104 can also receive communication signals (e.g., signal 124) from SV 112.

[0061] The wireless communication system 100 may also include one or more UEs, such as UE 190, which are indirectly connected to one or more communication networks via one or more device-to-device (D2D) or peer-to-peer (P2P) links (referred to as side links). Figure 1 In the example of FIG. 1 , UE 190 has a D2D P2P link 192 with one of UEs 104 connected to one of base stations 102 (e.g., through D2D P2P link 192, UE 190 can indirectly obtain cellular connection) and a D2D P2P link 194 with WLAN STA 152 connected to WLAN AP 150 (through D2D P2P link 194, UE 190 can indirectly obtain WLAN-based Internet connection). In one example, D2D P2P links 192 and 194 can be supported by any well-known D2D RAT, such as LTE Direct (LTE-D), WiFi Direct (WiFi-D), wait.

[0062] Figure 2A An example wireless network architecture 200 is shown. For example, the 5GC 210 (also referred to as the Next Generation Core "NGC") can be functionally viewed as a control plane (C-plane) function 214 (e.g., UE registration, authentication, network access, gateway selection, etc.) and a user plane (U-plane) function 212 (e.g., UE gateway function, access to data networks, IP routing, etc.), which operate in conjunction to form a core network. A user plane interface (NG-U) 213 and a control plane interface (NG-C) 215 connect the gNB 222 to the 5GC 210, specifically to the user plane function 212 and the control plane function 214, respectively. In an additional configuration, the ng-eNB 224 can also be connected to the 5GC 210 via the NG-C 215 to the control plane function 214 and the NG-U 213 to the user plane function 212. Further, the ng-eNB 224 can communicate directly with the gNB 222 via the backhaul connection 223. In some configurations, the next generation RAN (NG-RAN) 220 may have one or more gNBs 222, while other configurations include one or more of ng-eNBs 224 and gNBs 222. Either (or both) the gNB 222 or the ng-eNB 224 may communicate with one or more UEs 204 (e.g., any UE described herein).

[0063] Another optional aspect may include a location server 230 that can communicate with the 5GC 210 to provide location assistance for the UE 204. The location server 230 can be implemented as multiple independent servers (e.g., physically independent servers, different software modules on a single server, different software modules distributed on multiple physical servers, etc.), or alternatively, each location server 230 can correspond to a single server. The location server 230 can be configured to support one or more location services for the UE 204, and the UE 204 can be connected to the location server 230 via the core network, the 5GC 210, and / or via the Internet (not shown). Further, the location server 230 can be integrated into a component of the core network, or alternatively, can be external to the core network (e.g., a third-party server, such as an original equipment manufacturer (OEM) server or a business service server).

[0064] Figure 2B Another example wireless network structure 240 is shown. 5GC 260 (which may correspond to Figure 2AThe 5GC 210 in the UE 200 may be functionally viewed as a control plane function provided by an access and mobility management function (AMF) 264, and a user plane function provided by a user plane function (UPF) 262, which operate in conjunction to form a core network (i.e., 5GC 260). The functions of the AMF 264 include registration management, connection management, reachability management, mobility management, lawful interception, transmission of session management (SM) messages between one or more UEs 204 (any of the UEs described herein) and a session management function (SMF) 266, a transparent proxy service for routing SM messages, access authentication and access authorization, transmission of short message service (SMS) messages between the UE 204 and a short message service function (SMSF) (not shown), and a security anchor function (SEAF). The AMF 264 also interacts with an authentication server function (AUSF) (not shown) and the UE 204, and receives an intermediate key established as a result of the UE 204 authentication process. In the case of authentication based on the UMTS (Universal Mobile Telecommunications System) Subscriber Identity Module (USIM), the AMF 264 retrieves security material from the AUSF. The functions of the AMF 264 also include security context management (SCM). The SCM receives keys from the SEAF that it uses to derive access network-specific keys. The functions of the AMF 264 also include location service management for regulatory services, transmission of location service messages between the UE 204 and the Location Management Function (LMF) 270 (which can be used as the location server 230), transmission of location service messages between the NG-RAN 220 and the LMF 270, allocation of Evolved Packet System (EPS) bearer identifiers for interworking with EPS, and notification of UE 204 mobility events. In addition, the AMF 264 also supports the functions of non-3GPP (Third Generation Partnership Project) access networks.

[0065] The functions of UPF 262 include serving as an anchor point for intra-RAT / inter-RAT mobility (when applicable), serving as an external protocol data unit (PDU) session point for interconnection to a data network (not shown), providing packet routing and forwarding, packet inspection, user plane policy rule enforcement (e.g., gating, redirection, traffic manipulation), lawful interception (user plane collection), traffic usage reporting, user plane quality of service (QoS) processing (e.g., uplink / downlink rate enforcement, reflective QoS marking in downlink), uplink traffic verification (mapping of service data flow (SDF) to QoS flow), transport level packet marking in uplink and downlink, downlink packet buffering and downlink data notification triggering, and transmitting and forwarding one or more "end markers" to the source RAN node. UPF 262 may also support the transmission of location service messages on the user plane between UE 204 and a location server (such as SLP 272).

[0066] The functions of SMF 266 include session management, UE Internet Protocol (IP) address allocation and management, selection and control of user plane functions, configuring traffic manipulation in UPF 262 to route traffic to the correct destination, controlling partial policy enforcement and QoS, and downlink data notification. The interface through which SMF 266 communicates with AMF 264 is called the N11 interface.

[0067] Another optional aspect may include an LMF 270 that can communicate with the 5GC 260 to provide location assistance for the UE 204. The LMF 270 can be implemented as multiple independent servers (e.g., physically independent servers, different software modules on a single server, different software modules distributed on multiple physical servers, etc.), or alternatively, each LMF 270 can correspond to a single server. The LMF 270 can be configured to support one or more location services for the UE 204, and the UE 204 can connect to the LMF 270 via the core network, the 5GC 260, and / or via the Internet (not shown). The SLP 272 may support similar functionality as the LMF 270, but the LMF 270 may communicate with the AMF 264, the NG-RAN 220, and the UE 204 via a control plane (e.g., using interfaces and protocols intended to carry signaling messages rather than voice or data), whereas the SLP 272 may communicate with the UE 204 and external clients (third party servers 274) via a user plane (e.g., using protocols intended to carry voice or data, such as the Transmission Control Protocol (TCP) and / or IP).

[0068] Yet another optional aspect may include a third-party server 274 that may communicate with the LMF 270, SLP 272, 5GC 260 (e.g., via the AMF 264 and / or UPF 262), NG-RAN 220, and / or UE 204 to obtain location information (e.g., a location estimate) of the UE 204. As such, in some cases, the third-party server 274 may be referred to as a location service (LCS) client or external client. The third-party server 274 may be implemented as a plurality of independent servers (e.g., physically independent servers, different software modules on a single server, different software modules distributed across multiple physical servers, etc.), or alternatively, each location server 274 may correspond to a single server.

[0069] The user plane interface 263 and the control plane interface 265 connect the 5GC 260, specifically the UPF 262 and the AMF 264, respectively, to one or more gNBs 222 and / or ng-eNBs 224 in the NG-RAN 220. The interface between the gNB 222 and / or ng-eNB 224 and the AMF 264 is referred to as the "N2" interface, and the interface between the gNB 222 and / or ng-eNB 224 and the UPF 262 is referred to as the "N3" interface. The gNBs 222 and / or ng-eNBs 224 of the NG-RAN 220 may communicate directly with each other via the backhaul connection 223 (referred to as the "Xn-C" interface). One or more of the gNBs 222 and / or ng-eNBs 224 may communicate with one or more UEs 204 via a wireless interface referred to as the "Uu" interface.

[0070] The functions of the gNB 222 may be divided between a gNB central unit (gNB-CU) 226, one or more gNB distributed units (gNB-DU) 228, and one or more gNB radio units (gNB-RU) 229. The gNB-CU 226 is a logical node that includes base station functions such as transmission of user data, mobility control, radio access network sharing, positioning, session management, etc., except for those functions specifically assigned to the gNB-DU 228. More specifically, the gNB-CU 226 typically hosts the radio resource control (RRC), service data adaptation protocol (SDAP), and packet data convergence protocol (PDCP) protocols of the gNB 222. The gNB-DU 228 is a logical node that typically hosts the radio link control (RLC) and media access control (MAC) layers of the gNB 222. Its operation is controlled by the gNB-CU 226. One gNB-DU 228 may support one or more cells, and one cell is supported by only one gNB-DU 228. The interface 232 between the gNB-CU 226 and one or more gNB-DUs 228 is referred to as the "F1" interface. The physical (PHY) layer functions of the gNB 222 are typically hosted by one or more independent gNB-RUs 229 that perform functions such as power amplification and signal transmission / reception. The interface between the gNB-DU 228 and the gNB-RU 229 is referred to as the "Fx" interface. Thus, the UE 204 communicates with the gNB-CU 226 via the RRC, SDAP, and PDCP layers, with the gNB-DU 228 via the RLC and MAC layers, and with the gNB-RU 229 via the PHY layer.

[0071] The deployment of a communication system such as a 5G NR system can be arranged in a variety of ways with various components or constituent parts. 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, or a network device (such as a base station or one or more units (or one or more components) performing base station functions) can be implemented in an aggregated or decomposed architecture. For example, a base station (such as a Node B (NB), an evolved NB (eNB), a NR base station, a 5G NB, an access point (AP), a transmit receive point (TRP), or a cell, etc.) can be implemented as an aggregated base station (also called an independent base station or a monolithic base station) or a decomposed base station.

[0072] The aggregated base station can be configured to utilize a radio protocol stack that is physically or logically integrated within a single RAN node. The decomposed base station can be configured to utilize a protocol stack that is physically or logically distributed between two or more units, such as one or more central or centralized units (CUs), one or more distributed units (DUs), or one or more radio units (RUs). In some aspects, the CU can be implemented within a RAN node, and one or more DUs can be co-located with the CU, or alternatively, can be geographically or virtually distributed in one or more other RAN nodes. The DU can be implemented to communicate with one or more RUs. Each of the CU, DU, and RU can also be implemented as a virtual unit, namely a virtual central unit (VCU), a virtual distributed unit (VDU), or a virtual radio unit (VRU).

[0073] 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 used in an integrated access backhaul (IAB) network, an open radio access network (O-RAN (such as a network configuration sponsored by the O-RAN Alliance)), or a virtualized radio access network (vRAN, also known as a cloud radio access network (C-RAN)). Decomposition may include allocating functionality between two or more units at different physical locations, as well as virtually allocating functionality of at least one unit, which may enable flexibility in network design. The various units of a decomposed base station or decomposed RAN architecture may be configured for wired or wireless communication with at least one other unit.

[0074] Figure 2CAn example decomposed base station architecture 250 is shown in accordance with aspects of the present disclosure. The decomposed base station architecture 250 may include one or more central units (CUs) 280 (e.g., gNB-CUs 226), which may communicate directly with a core network 267 (e.g., 5GC 210, 5GC 260) via a backhaul link, or indirectly with a core network 267 through one or more decomposed base station units (e.g., a near real-time (near-RT) RAN intelligent controller (RIC) 259 via an E2 link, or a non-real-time (non-RT) RIC 257 associated with a service management and coordination (SMO) framework 255, or both). The CU 280 may communicate with one or more distributed units (DUs) 285 (e.g., gNB-DUs 228) via corresponding mid-haul links (e.g., F1 interfaces). The DUs 285 may communicate with one or more radio units (RUs) 287 (e.g., gNB-RUs 229) via corresponding fronthaul links. The RU 287 may communicate with a corresponding UE 204 via one or more radio frequency (RF) access links. In some implementations, a UE 204 may be served by multiple RUs 287 simultaneously.

[0075] Each unit, i.e., CU 280, DU 285, RU 287, as well as near-RT RIC 259, non-RT RIC 257, and SMO framework 255, 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 unit or an associated processor or controller that provides instructions to the communication interface of the unit may be configured to communicate with one or more of the other units via a transmission medium. For example, the units may include a wired interface configured to receive or send signals from one or more of the other units or to one or more of the other units via a wired transmission medium. Additionally, the units may include a wireless interface that may include a receiver, a transmitter, or a transceiver (such as a radio frequency (RF) transceiver) configured to receive or send signals or both receive and send signals to one or more of the other units via a wireless transmission medium.

[0076] In some aspects, CU 280 may host one or more higher layer control functions. Such control functions may include radio resource control (RRC), packet data convergence protocol (PDCP), service data adaptation protocol (SDAP), etc. Each control function may be implemented with an interface configured to transmit signals with other control functions hosted by CU 280. CU 280 may be configured to process user plane functions (i.e., central unit-user plane (CU-UP)), control plane functions (i.e., central unit-control plane (CU-CP)), or a combination thereof. In some embodiments, CU 280 may be logically divided into one or more CU-UP units and one or more CU-CP units. The CU-UP unit may communicate bidirectionally with the CU-CP unit via an interface (such as an E1 interface when implemented in an O-RAN configuration). As needed, CU 280 may be implemented to communicate with DU285 for network control and signaling.

[0077] DU 285 may correspond to a logical unit that includes one or more base station functions to control the operation of one or more RU 287. In some aspects, DU 285 may host one or more of a radio link control (RLC) layer, a media access control (MAC) layer, and one or more high physical (PHY) layers (such as modules for forward error correction (FEC) encoding and decoding, scrambling, modulation and demodulation, etc.) at least in part according to a functional partitioning such as that defined by the Third Generation Partnership Project (3GPP). In some aspects, DU 285 may also host one or more low PHY layers. Each layer (or module) may be implemented with an interface that is configured to transmit signals with other layers (and modules) hosted by DU 285 or with control functions hosted by CU 280.

[0078] The low-layer functions may be implemented by one or more RUs 287. In some deployments, a RU 287 controlled by a DU 285 may correspond to a logical node that hosts RF processing functions or low PHY layer functions (such as performing Fast Fourier Transform (FFT), Inverse FFT (iFFT), digital beamforming, physical random access channel (PRACH) extraction and filtering, etc.), or both, based at least in part on functional partitioning (such as low-layer functional partitioning). In such an architecture, the RU 287 may be implemented to handle over-the-air (OTA) communications with one or more UEs 204. In some embodiments, real-time and non-real-time aspects of control and user plane communications with the RU 287 may be controlled by the corresponding DU 285. In some cases, this configuration may enable the DU 285 and CU 280 to be implemented in a cloud-based RAN architecture, such as a vRAN architecture.

[0079] The SMO framework 255 may be configured to support RAN deployment and provisioning of non-virtualized and virtualized network elements. For non-virtualized network elements, the SMO framework 255 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 255 may be configured to interact with a cloud computing platform (such as an open cloud (O-Cloud) 269) 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 280, DU 285, RU 287, and near-RT RIC 259. In some embodiments, the SMO framework 255 may communicate with hardware aspects of the 4G RAN (such as an open eNB (O-eNB) 261) via the O1 interface. Additionally, in some embodiments, the SMO framework 255 may communicate directly with one or more RUs 287 via the O1 interface. The SMO framework 255 may also include a non-RT RIC 257 configured to support the functionality of the SMO framework 255 .

[0080] The non-RT RIC 257 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 259. The non-RT RIC 257 may be coupled to or in communication with the near-RT RIC 259 (such as via an A1 interface). The near-RT RIC 259 may be configured to include logic functions that implement near-real-time control and optimization of RAN elements and resources through data collection and actions on an interface (such as via an E2 interface) that connects one or more CUs 280, one or more DUs 285, or both, and the O-eNB with the near-RT RIC 259.

[0081] In some embodiments, in order to generate an AI / ML model to be deployed in the near-RT RIC 259, the non-RT RIC 257 may receive parameters or external enrichment information from an external server. Such information may be used by the near-RT RIC 259 and may be received from a non-network data source or from a network function at the SMO framework 255 or the non-RT RIC 257. In some examples, the non-RT RIC 257 or the near-RT RIC 259 may be configured to adjust RAN behavior or performance. For example, the non-RT RIC 257 may monitor long-term trends and patterns of performance and employ AI / ML models to perform corrective actions through the SMO framework 255 (such as via reconfiguration of O1) or via the creation of RAN management policies (such as A1 policies).

[0082] Figure 3A , Figure 3B and Figure 3C Several example components (represented by corresponding blocks) are shown, which may be incorporated into UE 302 (which may correspond to any UE described herein), base station 304 (which may correspond to any base station described herein), and network entity 306 (which may correspond to or implement any network function described herein, including location server 230 and NG-LMF 270, or alternatively, may be independent of Figure 2A and Figure 2B The NG-RAN 220 and / or 5GC 210 / 260 infrastructure shown may include a wireless communication network (such as a dedicated network) to support the operations described herein. It will be understood that these components may be implemented in different types of devices in different embodiments (e.g., in an ASIC, in a system on a chip (SoC), etc.). The components shown may also be incorporated into other devices in the communication system. For example, other devices in the system may include components similar to the components described to provide similar functionality. Similarly, a given device may include one or more of these components. For example, a device may include multiple transceiver components that enable the device to operate on multiple carriers and / or communicate via different technologies.

[0083] UE 302 and base station 304 each include one or more wireless wide area network (WWAN) transceivers 310 and 350, respectively, providing means (e.g., means for transmitting, means for receiving, means for measuring, means for tuning, means for suppressing transmission, etc.) for communicating via one or more wireless communication networks (not shown), such as NR networks, LTE networks, GSM networks, etc. WWAN transceivers 310 and 350 may each be connected to one or more antennas 316 and 356, respectively, for communicating with other network nodes (such as other UEs, access points, base stations (e.g., eNBs, gNBs)) via at least one designated RAT (e.g., NR, LTE, GSM, etc.) over a wireless communication medium of interest (e.g., a certain set of time / frequency resources in a particular spectrum). Depending on the specified RAT, the WWAN transceivers 310 and 350 can be configured differently to send and encode signals 318 and 358 (e.g., messages, indications, information, etc.), respectively, and conversely to receive and decode signals 318 and 358 (e.g., messages, indications, information, pilots, etc.), respectively. Specifically, the WWAN transceivers 310 and 350 include one or more transmitters 314 and 354 for sending and encoding signals 318 and 358, respectively, and one or more receivers 312 and 352 for receiving and decoding signals 318 and 358, respectively.

[0084] At least in some cases, the UE 302 and the base station 304 also include one or more short-range wireless transceivers 320 and 360, respectively. The short-range wireless transceivers 320 and 360 can be connected to one or more antennas 326 and 366, respectively, and provide for communicating over the wireless communication medium of interest via at least one dedicated RAT (e.g., WiFi, LTE-D, The short-range wireless transceivers 320 and 360 are components (e.g., components for sending, components for receiving, components for measuring, components for tuning, components for suppressing sending, etc.) for communicating with other network nodes (such as other UEs, access points, base stations, etc.) using a PC5, dedicated short-range communication (DSRC), wireless access for vehicular environments (WAVE), near field communication (NFC), ultra-wideband (UWB), etc.). Depending on the specified RAT, the short-range wireless transceivers 320 and 360 can be configured differently to send and encode signals 328 and 368 (e.g., messages, indications, information, etc.), respectively, and conversely configured to receive and decode signals 328 and 368 (e.g., messages, indications, information, pilots, etc.), respectively. Specifically, the short-range wireless transceivers 320 and 360 respectively include one or more transmitters 324 and 364 for sending and encoding signals 328 and 368, respectively, and one or more receivers 322 and 362 for receiving and decoding signals 328 and 368, respectively. As a specific example, the short-range wireless transceivers 320 and 360 may be WiFi transceivers, Transceiver, and / or NFC transceiver, UWB transceiver, or vehicle-to-vehicle (V2V) and / or vehicle-to-everything (V2X) transceiver.

[0085] At least in some cases, the UE 302 and the base station 304 also include satellite signal receivers 330 and 370. The satellite signal receivers 330 and 370 can be connected to one or more antennas 336 and 376, respectively, and can provide components for receiving and / or measuring satellite positioning / communication signals 338 and 378, respectively. In the case where the satellite signal receivers 330 and 370 are satellite positioning system receivers, the satellite positioning / communication signals 338 and 378 can be global navigation satellite system (GNSS) signals, global navigation satellite system (GLONASS) signals, Galileo signals, Beidou signals, Indian Regional Navigation Satellite System (NAVIC) signals, Quasi-Zenith Satellite System (QZSS) signals, GPS signals, etc. In the case where the satellite signal receivers 330 and 370 are non-terrestrial network (NTN) receivers, the satellite positioning / communication signals 338 and 378 can be communication signals (e.g., carrying control and / or user data) originating from a 5G network. Satellite signal receivers 330 and 370 may include any suitable hardware and / or software for receiving and processing satellite positioning / communication signals 338 and 378, respectively. Satellite signal receivers 330 and 370 may request appropriate information and operations from other systems and, at least in some cases, perform calculations using measurements obtained through any suitable satellite positioning system algorithm to determine the location of UE 302 and base station 304, respectively.

[0086] The base station 304 and the network entity 306 each include one or more network transceivers 380 and 390, respectively, to provide means (e.g., means for transmitting, means for receiving, etc.) for communicating with other network entities (e.g., other base stations 304, other network entities 306). For example, the base station 304 may use one or more network transceivers 380 to communicate with other base stations 304 or network entities 306 over one or more wired or wireless backhaul links. As another example, the network entity 306 may use one or more network transceivers 390 to communicate with one or more base stations 304 over one or more wired or wireless backhaul links, or communicate with other network entities 306 over one or more wired or wireless core network interfaces.

[0087] The transceiver can be configured to communicate over a wired or wireless link. The transceiver (whether a wired transceiver or a wireless transceiver) includes a transmitter circuit (e.g., transmitter 314, 324, 354, 364) and a receiver circuit (e.g., receiver 312, 322, 352, 362). In some embodiments, the transceiver can be an integrated device (e.g., implementing the transmitter circuit and the receiver circuit in a single device), in some embodiments, the transceiver can include a separate transmitter circuit and a separate receiver circuit, or in other embodiments, the transceiver can be implemented in other ways. The transmitter circuit and the receiver circuit of a wired transceiver (e.g., network transceivers 380 and 390 in some embodiments) can be coupled to one or more wired network interface ports. The wireless transmitter circuitry (e.g., transmitters 314, 324, 354, 364) may include or be coupled to multiple antennas (e.g., antennas 316, 326, 356, 366), such as antenna arrays, which allow the corresponding device (e.g., UE 302, base station 304) to perform transmit "beamforming", as described herein. Similarly, the wireless receiver circuitry (e.g., receiver 312, 322, 352, 362) may include or be coupled to multiple antennas (e.g., antennas 316, 326, 356, 366), such as antenna arrays, which allow the corresponding device (e.g., UE 302, base station 304) to perform receive beamforming, as described herein. In one aspect, the transmitter circuitry and the receiver circuitry may share the same multiple antennas (e.g., antennas 316, 326, 356, 366) so that the corresponding device can only receive or transmit at a given time, but not simultaneously. The wireless transceivers (eg, WWAN transceivers 310 and 350, short-range wireless transceivers 320 and 360) may also include a network listening module (NLM) or the like for performing various measurements.

[0088] As used herein, various wireless transceivers (e.g., transceivers 310, 320, 350, and 360 and network transceivers 380 and 390 in some embodiments) and wired transceivers (e.g., network transceivers 380 and 390 in some embodiments) may be generally characterized as a "transceiver," "at least one transceiver," or "one or more transceivers." As such, whether a particular transceiver is a wired or wireless transceiver may be inferred from the type of communication being performed. For example, backhaul communications between network devices or servers typically involve signaling via a wired transceiver, while wireless communications between a UE (e.g., UE 302) and a base station (e.g., base station 304) typically involve signaling via a wireless transceiver.

[0089] UE 302, base station 304, and network entity 306 also include other components that can be used in conjunction with the operations disclosed herein. UE 302, base station 304, and network entity 306 include one or more processors 332, 384, and 394, respectively, for providing functions related to, for example, wireless communication, and for providing other processing functions. Thus, processors 332, 384, and 394 can provide components for processing, such as components for determining, components for calculating, components for receiving, components for sending, components for indicating, etc. In one aspect, processors 332, 384, and 394 can include, for example, one or more general-purpose processors, multi-core processors, central processing units (CPUs), ASICs, digital signal processors (DSPs), field programmable gate arrays (FPGAs), other programmable logic devices or processing circuits, or various combinations thereof.

[0090] UE 302, base station 304, and network entity 306 include memory circuitry implementing memory 340, 386, and 396 (e.g., each including a memory device) for maintaining information (e.g., information indicating reserved resources, thresholds, parameters, etc.). Thus, memory 340, 386, and 396 may provide means for storing, means for retrieving, means for maintaining, etc. In some cases, UE 302, base station 304, and network entity 306 may include positioning components 342, 388, and 398, respectively. Positioning components 342, 388, and 398 may be hardware circuits that are part of or coupled to processors 332, 384, and 394, respectively, which, when executed, cause UE 302, base station 304, and network entity 306 to perform the functions described herein. In other aspects, the positioning components 342, 388, and 398 can be external to the processors 332, 384, and 394 (e.g., part of a modem processing system, integrated with another processing system, etc.). Alternatively, the positioning components 342, 388, and 398 can be memory modules stored in the memories 340, 386, and 396, respectively, which, when executed by the processors 332, 384, and 394 (or a modem processing system, another processing system, etc.), cause the UE 302, the base station 304, and the network entity 306 to perform the functions described herein. Figure 3A Possible locations are shown for location component 342, which can be, for example, part of one or more WWAN transceivers 310, memory 340, one or more processors 332, or any combination thereof, or can be a standalone component. Figure 3B Possible locations are shown for the positioning component 388, which can be, for example, part of one or more WWAN transceivers 350, the memory 386, the one or more processors 384, or any combination thereof, or can be a stand-alone component. Figure 3C Possible locations are shown for a location component 398, which may be, for example, part of one or more network transceivers 390, memory 396, one or more processors 394, or any combination thereof, or may be a stand-alone component.

[0091] UE 302 may include one or more sensors 344 coupled to one or more processors 332 to provide components for sensing or detecting movement and / or direction information independent of motion data derived from signals received by one or more WWAN transceivers 310, one or more short-range wireless transceivers 320, and / or satellite signal receivers 330. For example, sensor 344 may include an accelerometer (e.g., a micro-electromechanical system (MEMS) device), a gyroscope, a geomagnetic sensor (e.g., a compass), an altimeter (e.g., a barometric altimeter), and / or any other type of motion detection sensor. In addition, sensor 344 may include multiple different types of devices and combine their outputs to provide motion information. For example, sensor 344 may use a combination of a multi-axis accelerometer and an orientation sensor to provide the ability to calculate position in a two-dimensional (2D) and / or three-dimensional (3D) coordinate system.

[0092] In addition, UE 302 includes a user interface 346, which provides means for providing indications to a user (e.g., audible and / or visual indications) and / or for receiving user input (e.g., when a user actuates a sensing device such as a keypad, touch screen, microphone, etc.). Although not shown, base station 304 and network entity 306 may also include a user interface.

[0093] Referring to the one or more processors 384 in more detail, in a downlink, IP packets from the network entity 306 may be provided to the processor 384. The one or more processors 384 may implement the functions of the RRC layer, the Packet Data Convergence Protocol (PDCP) layer, the Radio Link Control (RLC) layer, and the Medium Access Control (MAC) layer. One or more processors 384 provide RRC layer functions associated with broadcasting of system information (e.g., master information block (MIB), system information block (SIB)), RRC connection control (e.g., RRC connection paging, RRC connection establishment, RRC connection modification, and RRC connection release), inter-RAT mobility, and measurement configuration for UE measurement reporting; PDCP layer functions associated with header compression / decompression, security (encryption, decryption, integrity protection, integrity verification), and handover support functions; RLC layer functions associated with transmission of upper layer PDUs, error correction through automatic repeat request (ARQ), concatenation, segmentation and reassembly of RLC service data units (SDUs), re-segmentation of RLC data PDUs, and reordering of RLC data PDUs; and MAC layer functions associated with mapping between logical channels and transport channels, scheduling information reporting, error correction, priority handling, and logical channel prioritization.

[0094] The transmitter 354 and the receiver 352 may implement layer-1 (L1) functions associated with various signal processing functions. Layer-1, including the physical (PHY) layer, may include error detection on the transmission channel, forward error correction (FEC) encoding / decoding of the transmission channel, interleaving, rate matching, mapping to the physical channel, modulation / demodulation of the physical channel, and MIMO antenna processing. The transmitter 354 processes the mapping to the signal constellation based on various modulation schemes (e.g., binary phase shift keying (BPSK), quadrature phase shift keying (QPSK), M-phase shift keying (M-PSK), M-quadrature amplitude modulation (M-QAM)). The coded and modulated symbols may then be divided into parallel streams. Each stream may then be mapped to an orthogonal frequency division multiplexing (OFDM) subcarrier, multiplexed with a reference signal (e.g., pilot) in the time domain and / or frequency domain, and then combined together using an inverse fast Fourier transform (IFFT) to produce a physical channel carrying a time domain OFDM symbol stream. The OFDM symbol stream is spatially precoded to produce multiple spatial streams. The channel estimate from the channel estimator can be used to determine the coding and modulation scheme, as well as for spatial processing. The channel estimate can be derived from a reference signal and / or channel condition feedback sent by the UE 302. Each spatial stream can then be provided to one or more different antennas 356. The transmitter 354 can modulate the RF carrier with the corresponding spatial stream for transmission.

[0095] At the UE 302, the receiver 312 receives the signal through its corresponding antenna 316. The receiver 312 recovers the information modulated onto the RF carrier and provides the information to one or more processors 332. The transmitter 314 and the receiver 312 implement layer-1 functions associated with various signal processing functions. The receiver 312 can perform spatial processing on the information to recover any spatial streams destined for the UE 302. If multiple spatial streams are destined for the UE 302, they can be combined into a single OFDM symbol stream by the receiver 312. The receiver 312 then converts the OFDM symbol stream from the time domain to the frequency domain using a fast Fourier transform (FFT). The frequency domain signal includes a separate OFDM symbol stream for each subcarrier of the OFDM signal. By determining the most likely signal constellation point sent by the base station 304, the symbols and reference signals on each subcarrier are recovered and demodulated. These soft decisions can be based on channel estimates calculated by a channel estimator. The soft decisions are then decoded and deinterleaved to recover the data and control signals originally sent by the base station 304 on the physical channel. The data and control signals are then provided to one or more processors 332, which implement layer-3 (L3) and layer-2 (L2) functions.

[0096] In the uplink, one or more processors 332 provide demultiplexing between transport and logical channels, packet reassembly, decryption, header decompression, and control signal processing to recover IP packets from the core network. One or more processors 332 are also responsible for error detection.

[0097] Similar to the functions described in conjunction with the downlink transmission of the base station 304, one or more processors 332 provide RRC layer functions associated with system information (e.g., MIB, SIB) acquisition, RRC connection, and measurement reporting; PDCP layer functions associated with header compression / decompression, security (encryption, decryption, integrity protection, integrity verification); RLC layer functions associated with transmission of upper layer PDUs, error correction through ARQ, concatenation, segmentation and reassembly of RLC SDUs, resegmentation of RLC data PDUs, and reordering of RLC data PDUs; and MAC layer functions associated with mapping between logical channels and transport channels, multiplexing of MAC SDUs onto transport blocks (TBs), demultiplexing of MAC SDUs from TBs, scheduling information reporting, error correction through hybrid automatic repeat request (HARQ), priority processing, and logical channel prioritization.

[0098] The transmitter 314 may select an appropriate coding and modulation scheme and facilitate spatial processing using channel estimates derived by a channel estimator from a reference signal or feedback sent by the base station 304. The spatial streams generated by the transmitter 314 may be provided to different antennas 316. The transmitter 314 may modulate an RF carrier with a corresponding spatial stream for transmission.

[0099] The uplink transmission is processed at the base station 304 in a manner similar to that described in conjunction with the receiver functionality at the UE 302. The receiver 352 receives the signal through its respective antenna 356. The receiver 352 recovers the information modulated onto the RF carrier and provides the information to one or more processors 384.

[0100] In the uplink, one or more processors 384 provide demultiplexing between transport and logical channels, packet reassembly, decryption, header decompression, control signal processing to recover IP packets from UE 302. IP packets from one or more processors 384 may be provided to the core network. One or more processors 384 are also responsible for error detection.

[0101] For convenience, UE 302, base station 304 and / or network entity 306 may Figure 3A , Figure 3B and Figure 3C1 is shown as including various components that can be configured according to the various examples described herein. However, it will be understood that in different designs, the components shown may have different functions. Specifically, FIG. 3A to FIG. 3C Various components in are optional in alternative configurations, and various aspects include configurations that may vary due to design choice, cost, use of the device, or other considerations. For example, in Figure 3A In the case of , a particular implementation of UE 302 may omit WWAN transceiver 310 (e.g., a wearable device or tablet or PC or laptop may have Wi-Fi and / or Bluetooth capabilities but no cellular capabilities), or may omit short-range wireless transceiver 320 (e.g., only cellular, etc.), or may omit satellite signal receiver 330, or may omit sensor 344, etc. In another example, in Figure 3B In the case of a wireless network, a specific implementation of the base station 304 may omit the WWAN transceiver 350 (e.g., a Wi-Fi "hotspot" access point without cellular capabilities), or may omit the short-range wireless transceiver 360 (e.g., cellular only, etc.), or may omit the satellite signal receiver 370, etc. For the sake of brevity, illustrations of various alternative configurations are not provided herein, but are readily apparent to those skilled in the art.

[0102] The various components of the UE 302, base station 304, and network entity 306 may be communicatively coupled to each other via data buses 334, 382, ​​and 392, respectively. In one aspect, the data buses 334, 382, ​​and 392 may form or be a part of a communication interface for the UE 302, base station 304, and network entity 306, respectively. For example, in the case where different logical entities are included in the same device (e.g., gNB and location server functions are merged into the same base station 304), the data buses 334, 382, ​​and 392 may provide communication between them.

[0103] Figure 3A , Figure 3B and Figure 3C The components can be implemented in various ways. In some embodiments, Figure 3A , Figure 3B and Figure 3CThe components may be implemented in one or more circuits, such as one or more processors and / or one or more ASICs (which may include one or more processors). Here, each circuit may use and / or combine at least one memory component to store information or executable code used by the circuit to provide the functionality. For example, some or all of the functions represented by blocks 310 to 346 may be implemented by the processor and memory components of UE 302 (e.g., by executing appropriate codes and / or by appropriate configuration of processor components). Similarly, some or all of the functions represented by blocks 350 to 388 may be implemented by the processor and memory components of base station 304 (e.g., by executing appropriate codes and / or by appropriate configuration of processor components). In addition, some or all of the functions represented by blocks 390 to 398 may be implemented by the processor and memory components of network entity 306 (e.g., by executing appropriate codes and / or by appropriate configuration of processor components). For simplicity, various operations, actions and / or functions are described herein as being performed by "UE", "base station", "network entity", etc. However, it will be understood that such operations, actions and / or functions may actually be performed by specific components or combinations of components of the UE 302, base station 304, network entity 306, etc., such as processors 332, 384, 394, transceivers 310, 320, 350 and 360, memories 340, 386 and 396, positioning components 342, 388, 398, etc.

[0104] In some designs, the network entity 306 may be implemented as a core network component. In other designs, the network entity 306 may be operated independently of a network operator or cellular network infrastructure (e.g., NG RAN 220 and / or 5GC 210 / 260). For example, the network entity 306 may be a component of a dedicated network that may be configured to communicate with the UE 302 via the base station 304 or independently of the base station 304 (e.g., via a non-cellular communication link such as WiFi).

[0105] NR supports many cellular network-based positioning techniques, including downlink-based, uplink-based, and downlink and uplink-based positioning methods. Downlink-based positioning methods include observed time difference of arrival (OTDOA) in LTE, downlink time difference of arrival (DL-TDOA) in NR, and downlink angle of departure (DL-AoD) in NR. Figure 4Examples of various positioning methods according to various aspects of the present disclosure are shown. In the OTDOA or DL-TDOA positioning process shown by scenario 410, the UE measures the difference between the arrival time (ToA) of the reference signal (e.g., positioning reference signal (PRS)) received from the base station pair, referred to as the reference signal time difference (RSTD) or arrival time difference (TDOA) measurement, and reports them to the positioning entity. More specifically, the UE receives identifiers (IDs) of a reference base station (e.g., a serving base station) and multiple non-reference base stations in auxiliary data. The UE then measures the RSTD between the reference base station and each non-reference base station. Based on the known positions of the base stations involved and the RSTD measurements, a positioning entity (e.g., a UE for UE-based positioning or a location server for UE-assisted positioning) can estimate the position of the UE.

[0106] For DL-AoD positioning illustrated by scenario 420, the positioning entity uses measurement reports of received signal strength measurements of multiple downlink transmit beams from the UE to determine the angle between the UE and the transmitting base station. The positioning entity can then estimate the position of the UE based on the determined angle and the known location of the transmitting base station.

[0107] Uplink-based positioning methods include uplink time difference of arrival (UL-TDOA) and uplink angle of arrival (UL-AoA). UL-TDOA is similar to DL-TDOA, but is based on uplink reference signals (e.g., sounding reference signals (SRS)) sent by the UE to multiple base stations. Specifically, the UE sends one or more uplink reference signals measured by a reference base station and multiple non-reference base stations. Each base station then reports the reception time of the reference signal (called relative time of arrival (RTOA)) to a positioning entity (e.g., a location server) that knows the location and relative timing of the base stations involved. Based on the receive-receive (Rx-Rx) time difference between the reported RTOA of the reference base station and the reported RTOA of each non-reference base station, the known location of the base stations, and their known timing offsets, the positioning entity can use TDOA to estimate the position of the UE.

[0108] For UL-AoA positioning, one or more base stations measure the received signal strength of one or more uplink reference signals (e.g., SRS) received from the UE on one or more uplink receive beams. The positioning entity uses the signal strength measurements and the angle of the receive beam to determine the angle between the UE and the base station. Then, based on the determined angle and the known positioning of the base station, the positioning entity can estimate the position of the UE.

[0109] Downlink and uplink based positioning methods include enhanced cell identification (E-CID) positioning and multiple round trip time (RTT) positioning (also known as "multi-cell RTT" and "multi-RTT"). In the RTT process, a first entity (e.g., a base station or UE) sends a first RTT-related signal (e.g., a PRS or SRS) to a second entity (e.g., a UE or a base station), and the second entity sends a second RTT-related signal (e.g., an SRS or PRS) back to the first entity. Each entity measures the time difference between the arrival time (ToA) of the received RTT-related signal and the transmission time of the transmitted RTT-related signal. This time difference is called the receive-transmit (Rx-Tx) time difference. The Rx-Tx time difference measurement can be made or adjusted to include only the time difference between the nearest time slot boundary of the received signal and the nearest time slot boundary of the transmitted signal. The two entities may then transmit their Rx-Tx time difference measurements to a location server (e.g., LMF 270), which calculates the round-trip propagation time (i.e., RTT) between the two entities based on the two Rx-Tx time difference measurements (e.g., calculating the RTT as the sum of the two Rx-Tx time difference measurements). Alternatively, one entity may send its Rx-Tx time difference measurement to the other entity, which then calculates the RTT. The distance between the two entities may be determined based on the RTT and a known signal speed (e.g., the speed of light). For multi-RTT positioning shown in scenario 430, a first entity (e.g., a UE or a base station) performs an RTT positioning process with multiple second entities (e.g., multiple base stations or UEs) so that the location of the first entity can be determined based on the distance to the second entity and the known location of the second entity (e.g., using multi-point positioning). As shown in scenario 440, RTT and multi-RTT methods may be combined with other positioning techniques such as UL AoA and DL-AoD to improve positioning accuracy.

[0110] The E-CID positioning method is based on radio resource management (RRM) measurements. In E-CID, the UE reports the serving cell ID, timing advance (TA), and the identifiers, estimated timing, and signal strength of the detected neighboring base stations. The UE's position is then estimated based on this information and the known locations of the base stations.

[0111] To assist the positioning operation, a location server (e.g., location server 230, LMF 270, SLP 272) may provide assistance data to the UE. For example, the assistance data may include an identifier of a base station (or a cell / TRP of a base station) according to which a reference signal is measured, reference signal configuration parameters (e.g., the number of consecutive time slots including PRS, the periodicity of consecutive time slots including PRS, a muting sequence, a frequency hopping sequence, a reference signal identifier, a reference signal bandwidth, etc.), and / or other parameters applicable to a particular positioning method. Alternatively, the assistance data may originate directly from the base station itself (e.g., in an overhead message that is broadcast periodically, etc.). In some cases, the UE is able to detect neighboring network nodes on its own without using assistance data.

[0112] In the case of OTDOA or DL-TDOA positioning process, the auxiliary data may also include the expected RSTD value and the associated uncertainty or search window around the expected RSTD. In some cases, the value range of the expected RSTD may be + / -500 microseconds (μs). In some cases, when any resource used for positioning measurement is in FR1, the value range of the uncertainty of the expected RSTD may be + / -32μs. In other cases, when all resources used for positioning measurement are in FR2, the value range of the uncertainty of the expected RSTD may be + / -8μs.

[0113] The location estimate may be referred to by other names, such as a position estimate, location, position, position determination, etc. The location estimate may be geodetic and include coordinates (e.g., latitude, longitude, and possibly altitude), or may be citometric and include a street address, postal address, or some other verbal description of the location. The location estimate may also be defined relative to some other known location, or in absolute terms (e.g., using latitude, longitude, and possibly altitude). The location estimate may include expected errors or uncertainties (e.g., by including an area or volume within which the location is expected to be included with some specified or default confidence level).

[0114] Various frame structures may be used to support downlink and uplink transmissions between network nodes (eg, base stations and UEs). Figure 5 5 is a diagram illustrating an example frame structure according to aspects of the present disclosure. The frame structure may be a downlink or uplink frame structure. Other wireless communication technologies may have different frame structures and / or different channels.

[0115] LTE, and in some cases NR, use orthogonal frequency division multiplexing (OFDM) on the downlink and single carrier frequency division multiplexing (SC-FDM) on the uplink. However, unlike LTE, NR can also choose to use OFDM on the uplink. OFDM and SC-FDM divide the system bandwidth into multiple (K) orthogonal subcarriers, which are also commonly called tones, frequency bins, etc. Each subcarrier can be modulated with data. In general, modulation symbols are transmitted in the frequency domain using OFDM and in the time domain using SC-FDM. The spacing between adjacent subcarriers can be fixed, and the total number of subcarriers (K) can depend on the system bandwidth. For example, the subcarrier spacing can be 15 kilohertz (kHz), and the minimum resource allocation (resource block) can be 12 subcarriers (or 180kHz). Thus, for a system bandwidth of 1.25, 2.5, 5, 10, or 20 megahertz (MHz), the nominal fast Fourier transform (FFT) size may be equal to 128, 256, 512, 1024, or 2048, respectively. The system bandwidth may also be divided into subbands. For example, a subband may cover 1.08 MHz (i.e., 6 resource blocks), and for a system bandwidth of 1.25, 2.5, 5, 10, or 20 MHz, there may be 1, 2, 4, 8, or 16 subbands, respectively.

[0116] LTE supports a single parameter set (subcarrier spacing (SCS), symbol length, etc.). In contrast, NR can support multiple parameter sets (μ), for example, 15kHz (μ=0), 30kHz (μ=1), 60kHz (μ=2), 120kHz (μ=3), and 240kHz (μ=4) or larger subcarrier spacings are available. In each subcarrier spacing, there are 14 symbols per time slot. For 15kHz SCS (μ=0), there is one time slot per subframe, 10 time slots per frame, the time slot duration is 1 millisecond (ms), the symbol duration is 66.7 microseconds (μs), and the maximum nominal system bandwidth (in MHz) with a 4K FFT size is 50. For 30kHz SCS (μ=1), there are two time slots per subframe, 20 time slots per frame, the time slot duration is 0.5ms, the symbol duration is 33.3μs, and the maximum nominal system bandwidth (in MHz) with a 4K FFT size is 100. For 60kHz SCS (μ=2), there are 4 slots per subframe, 40 slots per frame, the slot duration is 0.25ms, the symbol duration is 16.7μs, and the maximum nominal system bandwidth (in MHz) with a 4K FFT size is 200. For 120kHz SCS (μ=3), there are 8 slots per subframe, 80 slots per frame, the slot duration is 0.125ms, the symbol duration is 8.33μs, and the maximum nominal system bandwidth (in MHz) with a 4K FFT size is 400. For 240kHz SCS (μ=4), there are 16 slots per subframe, 160 slots per frame, the slot duration is 0.0625ms, the symbol duration is 4.17μs, and the maximum nominal system bandwidth (in MHz) with a 4K FFT size is 800.

[0117] exist Figure 5 In the example, a 15kHz parameter set is used. Therefore, in the time domain, the 10ms frame is divided into 10 equally sized subframes, each 1ms, and each subframe includes one time slot. Figure 5 , the horizontal direction (on the X-axis) represents time, which increases from left to right, and the vertical direction (on the Y-axis) represents frequency, which increases (or decreases) from bottom to top.

[0118] A resource grid can be used to represent a time slot, each of which includes one or more time-concurrent resource blocks (RBs) (also called physical RBs (PRBs)) in the frequency domain. The resource grid is further divided into a number of resource elements (REs). An RE can correspond to one symbol length in the time domain and one subcarrier in the frequency domain. Figure 5In the parameter set of , for a normal cyclic prefix, an RB can contain 12 consecutive subcarriers in the frequency domain and 7 consecutive symbols in the time domain, for a total of 84 REs. For an extended cyclic prefix, an RB can contain 12 consecutive subcarriers in the frequency domain and six consecutive symbols in the time domain, for a total of 72 REs. The number of bits carried by each RE depends on the modulation scheme.

[0119] Some REs may carry reference (pilot) signals (RS). Reference signals may include positioning reference signals (PRS), tracking reference signals (TRS), phase tracking reference signals (PTRS), cell-specific reference signals (CRS), channel state information reference signals (CSI-RS), demodulation reference signals (DMRS), primary synchronization signals (PSS), secondary synchronization signals (SSS), synchronization signal blocks (SSB), sounding reference signals (SRS), etc., depending on whether the frame structure shown is for uplink or downlink communication. Figure 5 Example locations of REs carrying reference signals (labeled "R") are shown.

[0120] A set of resource elements (REs) used for transmission of PRS is referred to as a "PRS resource". A set of resource elements may span multiple PRBs in the frequency domain and "N" (such as 1 or more) consecutive symbols within a slot in the time domain. In a given OFDM symbol in the time domain, a PRS resource occupies consecutive PRBs in the frequency domain.

[0121] The transmission of PRS resources within a given PRB has a specific comb size (also called "comb density"). The comb size "N" represents the subcarrier spacing (or frequency / tone spacing) within each symbol of the PRS resource configuration. Specifically, for comb size "N", the PRS is transmitted in every Nth subcarrier of the symbol of the PRB. For example, for comb-4, for each symbol of the PRS resource configuration, the RE corresponding to every fourth subcarrier (such as subcarriers 0, 4, 8) is used to transmit the PRS of the PRS resource. Currently, DL-PRS supports comb sizes of comb-2, comb-4, comb-6, and comb-12. Figure 5 An example PRS resource configuration for comb-4 (which spans four symbols) is shown. That is, the location of the shaded REs (marked as "R") indicates a comb-4 PRS resource configuration.

[0122] Currently, DL-PRS resources can span 2, 4, 6 or 12 consecutive symbols within a slot with a full frequency domain staggered pattern. DL-PRS resources can be configured in flexible (FL) symbols of any configured higher layer downlink or slot. There is a constant energy per resource element (EPRE) for all REs of a given DL-PRS resource. The following are the inter-symbol frequency offsets for comb sizes of 2, 4, 6 and 12 over 2, 4, 6 and 12 symbols. 2-symbol comb-2: {0, 1}; 4-symbol comb-2: {0, 1, 0, 1}; 6-symbol comb-2: {0, 1, 0, 1, 0, 1}; 12-symbol comb-2: {0, 1, 0, 1, 0, 1, 0, 1, 0, 1}; 4-symbol comb-4: {0, 2, 1, 3} (as Figure 5 example); 12-symbol comb-4: {0, 2, 1, 3, 0, 2, 1, 3, 0, 2, 1, 3}; 6-symbol comb-6: {0, 3, 1, 4, 2, 5}; 12-symbol comb-6: {0, 3, 1, 4, 2, 5, 0, 3, 1, 4, 2, 5}; and 12-symbol comb-12: {0, 6, 3, 9, 1, 7, 4, 10, 2, 8, 5, 11}.

[0123] A "PRS resource set" is a set of PRS resources used for transmission of a PRS signal, wherein each PRS resource has a PRS resource ID. In addition, the PRS resources in a PRS resource set are associated with the same TRP. A PRS resource set is identified by a PRS resource set ID and is associated with a specific TRP (identified by a TRP ID). In addition, the PRS resources in a PRS resource set have the same periodicity, a common muting pattern configuration, and the same repetition factor (such as "PRS-ResourceRepetitionFactor") across time slots. The period is the time from the first repetition of the first PRS resource of the first PRS instance to the same first repetition of the same first PRS resource of the next PRS instance. The periodicity may have a length selected from 2^μ*{4, 5, 8, 10, 16, 20, 32, 40, 64, 80, 160, 320, 640, 1280, 2560, 5120, 10240} time slots, where μ=0, 1, 2, 3. The repetition factor may have a length selected from {1, 2, 4, 6, 8, 16, 32} slots.

[0124] The PRS resource ID in a PRS resource set is associated with a single beam (or beam ID) sent from a single TRP (wherein a TRP can send one or more beams). That is, each PRS resource in a PRS resource set can be sent on a different beam, so that a "PRS resource", or simply "resource", may also be referred to as a "beam". It should be noted that this has no effect on whether the UE knows the TRP and the beam on which the PRS is sent.

[0125] A "PRS instance" or "PRS opportunity" is an instance of a periodically repeating time window (such as a set of one or more consecutive time slots) in which a PRS is expected to be sent. A PRS opportunity may also be referred to as a "PRS positioning opportunity", "PRS positioning instance", "positioning opportunity", "positioning instance", "positioning repetition", or simply "opportunity", "instance" or "repetition".

[0126] A "positioning frequency layer" (also referred to as a "frequency layer") is a set of one or more PRS resource sets across one or more TRPs that have the same values ​​for certain parameters. Specifically, the set of PRS resource sets has the same subcarrier spacing and cyclic prefix (CP) type (meaning that all parameter sets supported for the physical downlink shared channel (PDSCH) are also supported for PRS), the same point A, the same downlink PRS bandwidth value, the same starting PRB (and center frequency), and the same comb size. The point A parameter takes the value of the parameter "ARFCN-ValueNR" (where "ARFCN" stands for "Absolute Radio Frequency Channel Number") and is an identifier / code that specifies a pair of physical radio channels for transmission and reception. The downlink PRS bandwidth can have a granularity of 4 PRBs, a minimum of 24 PRBs, and a maximum of 272 PRBs. Currently, up to four frequency layers have been defined, and each TRP of each frequency layer can be configured with up to two PRS resource sets.

[0127] The concept of frequency layer is somewhat similar to the concept of component carrier and bandwidth part (BWP), but the difference is that component carrier and BWP are used by one base station (or macro cell base station and small cell base station) to send data channels, while frequency layer is used by several (usually three or more) base stations to send PRS. When the UE transmits its positioning capabilities to the network, the UE can indicate the number of frequency layers it can support, such as during the LTE Positioning Protocol (LPP) session. For example, the UE can indicate whether it can support one or four positioning frequency layers.

[0128] It should be noted that the terms "positioning reference signal" and "PRS" may generally refer to a specific reference signal used for positioning in an NR or LTE system. However, as used herein, the terms "positioning reference signal" and "PRS" may also refer to any type of reference signal that can be used for positioning, such as but not limited to PRS, TRS, PTRS, CRS, CSI-RS, DMRS, PSS, SSS, SSB, SRS, UL-PRS, etc. defined in LTE and NR. In addition, the terms "positioning reference signal" and "PRS" may refer to a downlink, uplink or sidelink positioning reference signal, unless the context indicates otherwise. If the type of PRS needs to be further distinguished, the downlink positioning reference signal may be referred to as "DL-PRS", the uplink positioning reference signal (e.g., SRS, PTRS for positioning) may be referred to as "UL-PRS", and the sidelink positioning reference signal may be referred to as "SL-PRS". In addition, for signals (e.g., DMRS) that can be sent in the downlink, uplink and / or sidelink, "DL", "UL" or "SL" may be added before the signal to distinguish the direction. For example, "UL-DMRS" is different from "DL-DMRS".

[0129] Figure 6 600 is a diagram representing a channel estimate of a multipath channel between a receiver device (e.g., any UE or base station described herein) and a transmitter device (e.g., any other UE or base station described herein) according to aspects of the present disclosure. The channel estimate represents the strength of a radio frequency (RF) signal (e.g., a PRS) received through a multipath channel as a function of time delay and may be referred to as a channel energy response (CER), a channel impulse response (CIR), or a power delay profile (PDP) of the channel. Thus, the horizontal axis is in units of time (e.g., milliseconds) and the vertical axis is in units of signal strength (e.g., decibels). It should be noted that a multipath channel is a channel between a transmitter and a receiver on which the RF signal follows multiple paths or multipaths due to the transmission of the RF signal on multiple beams and / or the propagation characteristics of the RF signal (e.g., reflection, refraction, etc.).

[0130] exist Figure 6 In the example of , the receiver detects / measures multiple clusters (four clusters) of channel taps. Each channel tap represents the multipath followed by the RF signal between the transmitter and the receiver. That is, the channel taps represent the arrival of the RF signal on the multipath. Each cluster of channel taps represents that the corresponding multipath follows substantially the same path. Different clusters may exist due to the RF signal being transmitted on different transmit beams (and therefore at different angles), or due to the propagation characteristics of the RF signal (e.g., due to reflections that may follow different paths), or both.

[0131] The clusters of all channel taps for a given RF signal represent the multipath channels (or simply channels) between the transmitter and the receiver. Figure 6 In the channel shown, the receiver receives a first cluster of two RF signals at the channel tap at time T1, a second cluster of five RF signals at the channel tap at time T2, a third cluster of five RF signals at the channel tap at time T3, and a fourth cluster of four RF signals at the channel tap at time T4. Figure 6 In the example of , because the first cluster of RF signals at time T1 arrives first, it is assumed to correspond to RF signals transmitted on a transmit beam aligned with line-of-sight (LOS) or the shortest path. The third cluster at time T3 consists of the strongest RF signals and may correspond to, for example, RF signals transmitted on a transmit beam aligned with a non-line-of-sight (NLOS) path, etc. It should be noted that although Figure 6 Two to five clusters of channel taps are shown, but it will be understood that there may be more or fewer clusters of channel taps than shown.

[0132] Machine learning can be used to generate models that can be used to facilitate various aspects associated with data processing. One specific application of machine learning involves the generation of measurement models for processing positioning reference signals (e.g., PRS), such as feature extraction, reporting of reference signal measurements (e.g., selecting which extracted features to report), and the like.

[0133] Machine learning models are generally divided into supervised and unsupervised. Supervised models can also be subdivided into regression models or classification models. Supervised learning involves learning a function that maps inputs to outputs based on example input-output pairs. For example, given a training data set with two variables, age (input) and height (output), a supervised learning model can be generated to predict a person's height based on age. In a regression model, the output is continuous. An example of a regression model is linear regression, which simply tries to find the line that best fits the data. Extensions of linear regression include multiple linear regression (e.g., finding the plane of best fit) and polynomial regression (e.g., finding the curve of best fit).

[0134] Another example of a machine learning model is a decision tree model. In a decision tree model, a tree structure is defined using multiple nodes. Decisions are made from the root node at the top of the decision tree to the leaf nodes (i.e. nodes that have no further children) at the bottom of the decision tree. Generally, the more nodes there are in a decision tree model, the higher the accuracy of the decision.

[0135] Another example of a machine learning model is a decision forest. Random forest is an ensemble learning technique based on decision trees. Random forest involves creating multiple decision trees using a bootstrapped dataset of the original data, and randomly selecting a subset of variables at each step of the decision tree. The model then selects the mode of all predictions of each decision tree. By relying on a "majority wins" model, the risk of error from a single tree is reduced.

[0136] Another example of a machine learning model is a neural network (NN). A neural network is essentially a network of mathematical equations. A neural network accepts one or more input variables and, by passing through the network of equations, produces one or more output variables. In other words, a neural network receives an input vector and returns an output vector.

[0137] Figure 7 An example neural network 700 in accordance with aspects of the present disclosure is shown. The neural network 700 includes an input layer "i" that receives "n" input(s) (shown as "input 1", "input 2", and "input n"), one or more hidden layers (shown as hidden layers "h1", "h2", and "h3") for processing the inputs from the input layer, and an output layer "o" that provides "m" output(s) (labeled "output 1" and "output m"). The number of inputs "n", hidden layers "h", and outputs "m" can be the same or different. In some designs, the hidden layer "h" can include linear functions and / or activation functions, with the nodes of each successive hidden layer (shown as circles) processing the linear function and / or activation function according to the nodes of the previous hidden layer.

[0138] In a classification model, the output is discrete. An example of a classification model is logistic regression. Logistic regression is similar to linear regression, but is used to model the probability of a finite number of outcomes (usually two). Essentially, the logistic equation is created in such a way that the output values ​​can only be "0" and "1". Another example of a classification model is a support vector machine. For example, for two classes of data, a support vector machine will find a hyperplane or boundary between the two classes of data that maximizes the margin between the two classes. There are many planes that can separate the two classes, but only one plane can maximize the margin or distance between the classes. Another example of a classification model is naive Bayes, which is based on Bayes' theorem. Other examples of classification models include decision trees, random forests, and neural networks, which are similar to the examples described above, but the output is discrete rather than continuous.

[0139] Unlike supervised learning, unsupervised learning is used to infer and discover patterns from input data without reference to labeled results. Two examples of unsupervised learning models include clustering and dimensionality reduction.

[0140] Clustering is an unsupervised technique that involves grouping or clustering data points. Clustering is often used for customer segmentation, fraud detection, and document classification. Common clustering techniques include k-means clustering, hierarchical clustering, mean shift clustering, and density-based clustering. Dimensionality reduction is the process of reducing the number of random variables under consideration by obtaining a set of principal variables. In simpler terms, dimensionality reduction is the process of reducing the dimensionality of a feature set (or, in simpler terms, reducing the number of features). Most dimensionality reduction techniques can be categorized as feature elimination or feature extraction. An example of dimensionality reduction is called principal component analysis (PCA). In its simplest sense, PCA involves projecting higher dimensional data (e.g., three dimensions) into a smaller space (e.g., two dimensions). This results in a reduced dimensionality of the data (e.g., two dimensions instead of three) while maintaining all the original variables in the model.

[0141] Regardless of which machine learning model is used, at a high level, a machine learning module (e.g., implemented by a processing system such as processor 332, 384, or 394) can be configured to iteratively analyze training input data (e.g., measurements of reference signals to / from various target UEs) and associate the training input data with an output data set (e.g., a set of possible candidate locations for the various target UEs) so that the same output data set can be determined later when presented with similar input data (e.g., other target UEs from the same or similar locations).

[0142] NR supports positioning based on radio frequency fingerprint (RFFP), which is a positioning technology that uses RFFP captured by a mobile device to determine the location of a mobile device. RFFP can be a histogram of received signal strength indicator (RSSI), CER, CIR, PDP, or channel frequency response (CFR). RFFP can represent a single channel (e.g., PRS) received from a transmitter, all channels received from a specific transmitter, or all channels detectable at a receiver. The RFFP measured by a mobile device (e.g., UE) and the position of the transmitter associated with the measured RFFP (i.e., the transmitter that sends the RF signal measured by the mobile device to determine the RFFP) can be used to determine (e.g., triangulate) the position of the mobile device.

[0143] Machine learning positioning techniques have been demonstrated to provide superior positioning performance when compared to traditional positioning solutions. In machine learning-RFFP based positioning, a machine learning model (e.g., neural network 700) takes the RFFP of a downlink reference signal (e.g., PRS) as input and outputs a positioning measurement (e.g., ToA, RSTD) or a mobile device position corresponding to the input RFFP. The machine learning model (e.g., neural network 700) is trained using "ground truth" (i.e., known) positioning measurements or mobile device positions as reference (i.e., expected) outputs of the RFFP training set.

[0144] For example, a machine learning model can be trained to determine the RSTD measurement of a pair of TRPs from the RFFP of the PRS sent by the TRP. The reference output for training such a model would be the correct (i.e., ground truth) RSTD measurement of the location of the mobile device when the mobile device obtains the RFFP measurement of the PRS. The network (e.g., a location server) can determine the RSTD expected for the pair of TRPs based on the known location of the mobile device and the known location of the (measured) TRPs involved. The known location of the mobile device can be determined based on multiple reported RSTD measurements and / or any other measurements reported by the mobile device (e.g., GPS measurements).

[0145] Figure 8 800 is a diagram illustrating the use of a machine learning model for RFFP-based positioning in accordance with aspects of the present disclosure. Figure 8 In the example of FIG. 1 , during the “offline” phase, RFFPs (e.g., CER / CIR / CFR) captured by the mobile device are stored in a database. The database may be located at the mobile device or a network entity (e.g., a location server), and each RFFP may include a number of RFFPs captured by one or more transmitters (in Figure 8 1 to BS N)). For UE-based downlink RFFP (DL-RFFP) positioning, the network (e.g., location server) configures the base station to send a downlink reference signal (e.g., PRS) to the mobile device, and the RFFP is the CER / CIR / CFR of the configured downlink reference signal detected by the mobile device.

[0146] When the mobile device measures RFFP, each measured RFFP is compared with the known position of the mobile device (in Figure 8 The location of the mobile device may be known via another positioning technology, such as the one described above with reference to Figure 4 It should be noted that although Figure 8 The RFFP information for a single mobile device is shown, but it will be understood that the RFFP information for multiple mobile devices may be collected and stored in the database.

[0147] Based on the information captured during the offline phase, a machine learning model (e.g., neural network 700) is trained to predict the location of the mobile device based on the RFFP measured by the mobile device. More specifically, a training set of RFFP measurements is used as input to the machine learning model, and the known location of the mobile device when the RFFP is captured is used as a label. After training, during the "online" phase, the trained machine learning model can be used to predict (infer) the location of the mobile device (shown as "Pos M") based on the RFFP currently measured by the mobile device. For UE-based RFFP positioning, the network (e.g., a location server) provides the trained machine learning model to the mobile device. For UE-assisted positioning, the mobile device can provide RFFP measurements to the network for processing.

[0148] It should be noted that although Figure 8 The use of a machine learning model based on RFFP to estimate the position of the UE is shown, but the output (or extracted features) of the machine learning model can be changed to positioning measurements based on the input RFFP, such as RSTD measurements, ToA measurements, DL-AoD measurements, etc.

[0149] Fig. 9 900 is a diagram illustrating an inference cycle for UE-based DL-RFFP positioning according to aspects of the present disclosure. Fig. 9 As shown, the location server (e.g., LMF 270) configures DL-PRS resources so that they are sent by one or more TRPs during a positioning session with the UE. The TRP then sends the configured DL-PRS to the UE, and the UE measures the RFFP of the DL-PRS.

[0150] exist Fig. 9 In the example of FIG. 1 , the location server previously trained a machine learning model (labeled “RFFP ML”) for RFFP positioning, as shown in the above reference. Figure 7 and Figure 8 As discussed. The location server provides the machine learning model to the UE during the positioning session to perform inference (e.g., determine positioning measurements based on the measured RFFP). In this way, after measuring the RFFP of the DL-PRS, the UE inputs the measured RFFP into the received machine learning model to obtain the associated positioning measurements (e.g., ToA, RSTD).

[0151] Fig.10An example call flow 1000 for UE-based downlink-based RFFP positioning in accordance with aspects of the present disclosure is shown. In phase 1, the UE 204 and the LMF 270 perform an LPP positioning capability transfer procedure, during which the UE 204 provides its positioning capabilities to the LMF 270. In phase 2, the LMF 270 provides assistance information to the UE 204 of the serving ng-eNB / gNB 222 / 224 and any neighboring ng-eNB / gNB 222 / 224, such as the PRS resource configuration of the DL-PRS to be sent to the UE 204. In phase 3, the UE 204 and the LMF 270 perform an LPP assistance data exchange. During the exchange, the LMF 270 provides the UE 204 with assistance data for the positioning session, such as the configuration of the DL-PRS sent by the involved ng-eNB / gNB 222 / 224, and a machine learning model for reporting positioning measurements of the DL-PRS.

[0152] In phase 4, the LMF 270 optionally provides assistance information to the involved ng-eNB / gNB 222 / 224 via a New Radio Positioning Protocol Type A (NRPPa) message. In phase 5, the serving ng-eNB / gNB 222 / 224 optionally broadcasts the assistance information received from the LMF 270 in one or more positioning SIBs (posSIBs) as assistance data. In phase 6, the LMF 270 and the UE 204 perform an LPP Request / Provide Location Information procedure, during which the UE 204 provides positioning measurements on the DL-PRS sent by the ng-eNB / gNB 222 / 224. The positioning measurements can be derived by applying the machine learning model received in the assistance data to the RFFP of the measured DL-PRS.

[0153] Machine learning tools, their impact on the air interface, and their lifecycle management are currently being studied, using some representative use cases as a guide. As mentioned above, one of the use cases is positioning. The identified research areas include characterizing the lifecycle management of AI / ML models, such as model training, model deployment, model inference, model monitoring, and model updates. Research areas also include datasets for training, validation, testing, and inference.

[0154] RFFP is one of the promising data-driven positioning technologies. The goal of RFFP-based positioning is to develop AI / ML-based technologies for determining the UE location. However, training RFFP-based positioning models in practice is challenging due to the large amount of data required and the context-dependency of data-based learning algorithms. It should be noted that "RFFP" can be used as a general term to refer to any AI / ML positioning model.

[0155] One approach to addressing the challenges of RFFP training is through the concept of federated learning (FL). In federated learning, many devices perform local training themselves, and the local final weights are aggregated at a central server, which combines the local results and generates the final combined result (e.g., by averaging the generated local weights). In one aspect, the server that aggregates the weights can be implemented at an edge server, as an example application of mobile edge computing (MEC).

[0156] In the federated learning protocol, a client (e.g., UE) downloads a trainable model from a server (e.g., MEC 5G server). The client updates the model with its own data and then uploads the updated model to the server. The server aggregates multiple client updates to improve the model.

[0157] The present disclosure describes techniques for training node (e.g., UE) configuration and node selection for downlink-based RFFP in the context of federated learning. As used herein, the term "network" refers to a central entity that performs UE selection and aggregation phases in a federated learning protocol. The network can be an LMF 270, a model repository server, an edge server, etc. The focus of the present disclosure is on federated training based on downlink reference signals (e.g., PRS). It is assumed that the configuration of the training signal has been completed.

[0158] In the technology of the present disclosure, the network configures the UE with a RFFP-specific training configuration, which includes (1) an uncertainty threshold for the training samples to be used (e.g., the UE only uses training samples whose position uncertainty estimate is below a configurable threshold), and (2) a positioning method that the UE needs to follow to obtain its positioning (e.g., GNSS-based, cellular-based) to be used for training. It should be noted that in addition to the RFFP-specific features, the network can configure the UE with training features and parameters (e.g., mini-batch size, learning rate, optimization method, etc.).

[0159] Prior to node selection, the network retrieves information from configured UEs that can participate in training. The collected information is used for node selection. Each UE can indicate the size of the local data set it has collected, the quality of its downlink channel with the network (for sharing initial RFFP weights), and its current power budget. The network can estimate the quality of the uplink channel from each UE (for sharing updated weights). Based on the available nodes (i.e., UEs) and their available characteristics (e.g., data set size, channel quality, power), the network selects a group of nodes to participate in the RFFP federated learning training process. Discussed below are several criteria according to which the network can select nodes for training.

[0160] The first criterion is the regional identifier (ID) criterion. Here, the network selects UEs based on their geographical regions. For example, the network may have trained its RFFP model based on data from "region ID #1" and "region ID #2". Configured UEs whose datasets (at least most of them) belong to one of the trained regions (e.g., region ID #1) may not be selected by the network, while other UEs whose datasets cover other geographical regions may be favored in the selection, as this may enhance the training process and increase the robustness of the RFFP model to environmental changes.

[0161] The area ID may be based on an already defined and signaled identifier, such as a Tracking Area (TA) ID. Alternatively, the area ID may be based on a network implementation of area division. For example, the network may define the geographic boundaries of the area and identify which UEs to select based on (approximate) knowledge of the UE's geographic area.

[0162] The second criterion is the coverage area criterion. Here, the network selects the UE whose dataset covers the largest area. For example, two UEs may belong to the same area ID (as in the previous criteria), but one of the UEs may cover a larger area due to its mobility. Such UEs may be favored by the network. For example, Road Side Unit (RSU) equipment is expected to be fixed and may not provide the desired diversity for training purposes.

[0163] The third criterion is the local RFFP dataset size criterion. Here, the network selects UEs based on their dataset size. For example, if the network configures datasets for 10 UEs for RFFP federated learning training and only wants to select 5 UEs, the network selects the 5 UEs with the largest dataset size.

[0164] The fourth criterion is the RFFP training load balancing criterion. Here, the network selects UEs based on their previous participation level to ensure load balancing among multiple UEs. For example, if the network has two UEs with similar properties and data sets, it selects the UE that participated less in the training process. The training process can involve many rounds, and the network can track the identities of the UEs that contributed to the RFFP model training.

[0165] The fifth criterion is the UE training processing capability criterion. Here, the network selects UEs with better processing capabilities. Better processing capabilities allow the network to train the model faster.

[0166] The sixth criterion is the communication channel condition criterion for the downlink channel or the uplink channel. Here, the network can select a UE with a good downlink channel condition because the network needs to transmit the model weights to the UE to be updated. The network can also select a UE with a good uplink channel condition because the UE needs to upload the updated model weights to the network for aggregation. The network can select a UE with a good downlink channel condition and an uplink channel condition.

[0167] The seventh criterion is the RFFP test set performance criterion. Here, the network can select the UE with better RFFP test set performance.

[0168] The eighth criterion is any combination of the previous criteria. Here, the network can use a subset (or all) of the previous criteria to select UEs in the federated learning training phase.

[0169] In summary, in the technology described above, the process is as follows: First, the UE configured for data collection collects training data within a specific time period. Second, when training is performed on the UE side, the network requests information from the UE, such as area coverage, area ID, etc. Third, the network selects the UE for the training process and shares the model weights with the selected UE. Fourth, the selected UE performs the training phase and shares the updated model weights with the network. Fifth, the network aggregates the results from the selected UE and updates the model used for aggregation. The network can then use the model for RFFP-based positioning. More specifically, the network can provide the model to the UE for UE-based positioning, or apply the model to measurements reported by the UE for UE-assisted positioning.

[0170] The process described above may result in large storage overhead on the UE side. In addition, the UE that may have stored the training data may not be selected by the network, in which case the UE's resources are not used efficiently. In view of these considerations, the present disclosure proposes an alternative technique for a federated learning setting, in which only selected UEs collect training data and use it for training.

[0171] Specifically, training is divided into two phases: (1) a monitoring phase, and (2) a training phase. In the monitoring (or observation or sensing) phase, the UE monitors whether it meets the selection criteria set by the network. That is, the network can configure (or request) a certain set of UEs with the selection criteria to monitor within a certain time period and / or until configured / requested to provide monitoring results to the network. The selection criteria can be any of the criteria described above, and the network can indicate both the criteria and the expected value of the criteria. For example, the UE can be configured to monitor whether it is in a specific area ID, whether it is moving (i.e., not stationary), etc. If the criteria are met within a specified time period or when the network requests the monitoring results, the UE can (1) indicate to the network that the criteria are met (autonomously or upon request of the network), or (2) directly enter the training phase. Which option the UE should follow can be determined and configured by the network, and can be different for different UEs / UE sets.

[0172] In the training phase, the network shares the positioning model with the UE that has met the selection criteria (i.e., sends the model to the UE). It should be noted that this phase may not be required every time. For example, when the monitoring phase and the training phase are repeated multiple times in succession, the network can share the model at the beginning of the process, and the UE performs the monitoring and training phases multiple times.

[0173] During the training phase, each UE collects training data and updates its local model. The training data criteria are determined by the network, including the type of positioning data (e.g., measurements of CER, CFR, PDP, ToA, RSTD, etc.), conditions (e.g., the quality of the positioning estimate to be used, channel characteristics, etc.). Each UE then shares the updated model with the network (i.e., sends the model to the network). For example, sharing the model with the network can be performed (1) every N update phases (or iterations) or (2) at the end of a time window. Regarding the every N update iterations option, once the UE has performed N training steps, the UE reports the model weights and / or gradients to the network. The variable N can be specified by the network. Regarding the time window option, if the selection criteria are met, the UE updates the model with the collected training data during a specific time window configured by the network. The UE then reports the updated model weights and / or gradients to the network.

[0174] Fig.11 An example method 1100 of training a machine learning model according to aspects of the present disclosure is shown. In one aspect, the method 1100 may be performed by a UE (e.g., any UE described herein).

[0175] At 1110, the UE receives one or more selection criteria from a network entity (e.g., a server) for determining whether the UE is to participate in training a machine learning model. In one aspect, operation 1110 may be performed by one or more WWAN transceivers 310, one or more short-range wireless transceivers 320, one or more processors 332, memory 340, and / or positioning component 342, any or all of which may be considered as means for performing the operation.

[0176] At 1120, the UE determines whether the UE satisfies the one or more selection criteria (based on UE-specific values ​​collected / obtained by the UE) during the first time period. In one aspect, operation 1120 may be performed by one or more WWAN transceivers 310, one or more short-range wireless transceivers 320, one or more processors 332, memory 340, and / or positioning component 342, any or all of which may be considered means for performing the operation.

[0177] At 1130, after the second time period, the UE sends updated parameters of the machine learning model to the network entity, wherein the machine learning model is updated during the second time period based on a determination that the UE satisfies the one or more selection criteria. In one aspect, operation 1130 may be performed by one or more WWAN transceivers 310, one or more short-range wireless transceivers 320, one or more processors 332, memory 340, and / or positioning component 342, any or all of which may be considered means for performing the operation.

[0178] Fig.12 An example method 1200 of training a machine learning model according to aspects of the present disclosure is shown. In one aspect, the method 1200 may be performed by a network entity (e.g., a server).

[0179] At 1210, the network entity sends one or more selection criteria to a set of UEs (e.g., any UE described herein) for determining whether the set of UEs is to participate in training a machine learning model. In one aspect, operation 1210 may be performed by one or more network transceivers 390, one or more processors 394, memory 396, and / or positioning component 398, any or all of which may be considered means for performing the operation.

[0180] At 1220, the network entity sends the machine learning model to at least a subset of the UEs in the set of UEs that meet the one or more selection criteria. In one aspect, operation 1220 may be performed by one or more network transceivers 390, one or more processors 394, memory 396, and / or positioning component 398, any or all of which may be considered means for performing the operation.

[0181] At 1230, the network entity receives updated parameters of the machine learning model from each UE in the subset of UEs. In one aspect, operation 1230 may be performed by one or more network transceivers 390, one or more processors 394, memory 396, and / or positioning component 398, any or all of which may be considered means for performing the operation.

[0182] At 1240, the network entity updates the machine learning model based on the updated parameters received from each UE in the subset of UEs. In one aspect, operation 1240 may be performed by one or more network transceivers 390, one or more processors 394, memory 396, and / or positioning component 398, any or all of which may be considered means for performing the operation.

[0183] It should be understood that the technical advantage of methods 1100 and 1200 is that the efficiency of federated learning is improved because the UE does not store data that is not necessary for training the machine learning model. Another technical advantage is that the robustness of the RFFP training model is improved. Since less data is sent, the efficiency of over-the-air (OTA) resource utilization increases, but enough UEs send updated weights to the network to improve the robustness of the RFFP model.

[0184] In the above detailed description, it can be seen that different features are grouped together in the examples. This disclosure should not be understood as an example clause having more features than those explicitly mentioned in each clause. Instead, various aspects of the present disclosure may include less than all the features of the individual example clauses disclosed. Therefore, the following clauses should be considered to be included in the specification, where each clause itself can be used as a separate example. Although each dependent clause can refer to a specific combination of one of the other clauses in the clause, the aspect of the dependent clause is not limited to the specific combination. It should be understood that other example clauses may also include a combination of the subject matter of the dependent clause with any other dependent clause or independent clause, or any feature with other dependent and independent clauses. The various aspects disclosed herein explicitly include these combinations, unless it is explicitly expressed or it can be easily inferred that a specific combination is not intended (for example, contradictory aspects, such as defining an element as an electrical insulator and an electrical conductor). In addition, it is also intended that aspects of the clause can be included in any other independent clause, even if the clause is not directly dependent on the independent clause.

[0185] The following numbered clauses describe examples of implementation:

[0186] Clause 1. A method for training a machine learning model performed by a user equipment (UE), comprising: receiving one or more selection criteria from a network entity for determining whether the UE is to participate in training the machine learning model; determining whether the UE satisfies the one or more selection criteria during a first time period; and after a second time period, sending updated parameters of the machine learning model to the network entity, wherein the machine learning model is updated during the second time period based on a determination that the UE satisfies the one or more selection criteria.

[0187] Clause 2. The method of clause 1, further comprising: sending an indication to a network entity that the UE satisfies the one or more selection criteria.

[0188] Clause 3. The method according to any one of clauses 1 or 2 further includes: receiving a configuration from a network entity for reporting whether the UE satisfies the one or more selection criteria before the machine learning model is updated; or receiving a configuration from a network entity for updating the machine learning model based on a determination that the UE satisfies the one or more selection criteria and not reporting whether the UE satisfies the one or more selection criteria.

[0189] Clause 4. The method of any of clauses 1 to 3, further comprising: receiving a configuration of the first time period, the second time period, or both from a network entity.

[0190] Clause 5. The method according to any one of clauses 1 to 4 further includes: receiving a machine learning model from a network entity.

[0191] Clause 6. A method according to clause 5, wherein the machine learning model is received at the following time: before the first time period, or after the first time period and before the machine learning model is updated.

[0192] Clause 7. The method according to any one of clauses 1 to 6 also includes: performing multiple repetitions of determining whether the UE meets the one or more selection criteria and updating the machine learning model.

[0193] Clause 8. The method according to Clause 7 also includes: receiving a new machine learning model from a network entity for each of multiple repetitions; or receiving a single machine learning model from a network entity for multiple repetitions, wherein the machine learning model is a single machine learning model.

[0194] Clause 9. A method according to any one of clauses 1 to 8, wherein the machine learning model is trained based on training data collected by the UE during the second time period.

[0195] Clause 10. The method of clause 9, further comprising: receiving, from a network entity, a configuration of a type of training data to be collected.

[0196] Clause 11. A method according to any one of clauses 1 to 10, wherein the second time period includes: one or more update iterations of the machine learning model, or a time window.

[0197] Clause 12. A method according to any one of clauses 1 to 11, wherein the updated parameters include updated weights of the machine learning model, updated gradients of the machine learning model, or both.

[0198] Clause 13. A method according to any one of clauses 1 to 12, wherein the one or more selection criteria include: an area identifier criterion, a coverage area criterion, a local data set size criterion, a training load balancing criterion, a UE training processing capability criterion, a communication channel condition criterion, a test set performance criterion, or any combination thereof.

[0199] Clause 14. A method according to any one of clauses 1 to 13, wherein the machine learning model is a radio frequency fingerprint (RFFP) based machine learning model.

[0200] Clause 15. The method of any of clauses 1 to 14, wherein the network entity is a location server, an edge server, or a model repository server.

[0201] Clause 16. A method for training a machine learning model performed by a network entity, comprising: sending one or more selection criteria to a set of user equipment (UE) for determining whether the set of UEs is to participate in training the machine learning model; sending the machine learning model to at least one subset of UEs in the set of UEs that meet the one or more selection criteria; receiving updated parameters of the machine learning model from each UE in the subset of UEs; and updating the machine learning model based on the updated parameters received from each UE in the subset of UEs.

[0202] Clause 17. The method of clause 16, further comprising: receiving an indication from each UE in the subset of UEs that the UE satisfies the one or more selection criteria.

[0203] Clause 18. A method according to clause 17, wherein, in response to receiving an indication that the UE satisfies the one or more selection criteria, the machine learning model is sent to the UE.

[0204] Clause 19. The method according to any one of clauses 16 to 18 further includes: before training the machine learning model, sending a configuration to each UE in the UE set for reporting whether the UE satisfies the one or more selection criteria; or sending a configuration to each UE in the UE set for training the machine learning model based on a determination that the UE satisfies the one or more selection criteria and not reporting whether the UE satisfies the one or more selection criteria.

[0205] Clause 20. A method according to any one of clauses 16 to 19, further comprising: sending a configuration of a time period to each UE in the set of UEs, during which time period the value of the one or more selection criteria is monitored to determine whether the UE satisfies the one or more selection criteria.

[0206] Clause 21. The method according to any one of clauses 16 to 20 further includes: sending a configuration of a time period to each UE in the UE subset, during which the machine learning model is trained.

[0207] Clause 22. A method according to clause 21, wherein the machine learning model is trained based on training data collected by a subset of UEs during a time period.

[0208] Clause 23. The method of clause 22, further comprising: sending a configuration of the type of training data to be collected to each UE in the subset of UEs.

[0209] Clause 24. A method according to any one of clauses 21 to 23, wherein the time period includes: one or more update iterations of the machine learning model, or a time window.

[0210] Clause 25. A method according to any one of clauses 16 to 24, wherein the updated parameters include updated weights of the machine learning model, updated gradients of the machine learning model, or both.

[0211] Clause 26. A method according to any one of clauses 16 to 25, wherein the one or more selection criteria include: an area identifier criterion, a coverage area criterion, a local data set size criterion, a training load balancing criterion, a UE training processing capability criterion, a communication channel condition criterion, a test set performance criterion, or any combination thereof.

[0212] Clause 27. A method according to any one of clauses 16 to 26, wherein the machine learning model is a radio frequency fingerprint (RFFP) based machine learning model.

[0213] Clause 28. The method of any of clauses 16 to 27, wherein the network entity is a location server, an edge server, or a model repository server.

[0214] Clause 29. A user equipment (UE) comprising: a memory; at least one transceiver; and at least one processor communicatively coupled to the memory and the at least one transceiver, the at least one processor being configured to: receive one or more selection criteria for determining whether the UE is to participate in training a machine learning model from a network entity via the at least one transceiver; determine whether the UE satisfies the one or more selection criteria during a first time period; and after a second time period, send updated parameters of the machine learning model to the network entity via the at least one transceiver, wherein the machine learning model is updated during the second time period based on a determination that the UE satisfies the one or more selection criteria.

[0215] Clause 30. The UE of clause 29, wherein the at least one processor is further configured to: send an indication to a network entity via the at least one transceiver that the UE satisfies the one or more selection criteria.

[0216] Clause 31. A UE according to any one of clauses 29 to 30, wherein at least one processor is further configured to: receive, via at least one transceiver, a configuration for reporting whether the UE satisfies the one or more selection criteria from a network entity before the machine learning model is updated; or receive, via at least one transceiver, a configuration for updating the machine learning model based on a determination that the UE satisfies the one or more selection criteria and not reporting whether the UE satisfies the one or more selection criteria from a network entity.

[0217] Clause 32. The UE of any of clauses 29 to 31, wherein the at least one processor is further configured to: receive a configuration of the first time period, the second time period, or both from a network entity via the at least one transceiver.

[0218] Clause 33. A UE according to any one of clauses 29 to 32, wherein the at least one processor is further configured to: receive the machine learning model from a network entity via at least one transceiver.

[0219] Clause 34. A UE according to clause 33, wherein the machine learning model is received at the following time: before the first time period, or after the first time period and before the machine learning model is updated.

[0220] Clause 35. A UE according to any one of clauses 29 to 34, wherein at least one processor is further configured to: perform multiple repetitions of determining whether the UE satisfies the one or more selection criteria and updating the machine learning model.

[0221] Clause 36. A UE according to clause 35, wherein at least one processor is further configured to: receive a new machine learning model from a network entity via at least one transceiver for each of multiple repetitions; or receive a single machine learning model from a network entity via at least one transceiver for multiple repetitions, wherein the machine learning model is a single machine learning model.

[0222] Clause 37. A UE according to any one of clauses 29 to 36, wherein the machine learning model is trained based on training data collected by the UE during the second time period.

[0223] Clause 38. The UE of clause 37, wherein the at least one processor is further configured to: receive, via the at least one transceiver, from a network entity a configuration of a type of training data to be collected.

[0224] Clause 39. A UE according to any one of clauses 29 to 38, wherein the second time period comprises: one or more update iterations of the machine learning model, or a time window.

[0225] Clause 40. A UE according to any one of clauses 29 to 39, wherein the updated parameters include updated weights of the machine learning model, updated gradients of the machine learning model, or both.

[0226] Clause 41. A UE according to any one of clauses 29 to 40, wherein the one or more selection criteria include: an area identifier criterion, a coverage area criterion, a local data set size criterion, a training load balancing criterion, a UE training processing capability criterion, a communication channel condition criterion, a test set performance criterion, or any combination thereof.

[0227] Clause 42. A UE according to any one of clauses 29 to 41, wherein the machine learning model is a radio frequency fingerprint (RFFP) based machine learning model.

[0228] Clause 43. The UE of any of clauses 29 to 42, wherein the network entity is a location server, an edge server, or a model repository server.

[0229] Clause 44. A network entity comprising: a memory; at least one transceiver; and at least one processor communicatively coupled to the memory and the at least one transceiver, the at least one processor being configured to: send one or more selection criteria for determining whether the set of UEs is to participate in training a machine learning model to a set of user equipment (UEs) via at least one transceiver; send the machine learning model to at least one subset of UEs in the set of UEs that meet the one or more selection criteria via at least one transceiver; receive updated parameters of the machine learning model from each UE in the subset of UEs via at least one transceiver; and update the machine learning model based on the updated parameters received from each UE in the subset of UEs.

[0230] Clause 45. The network entity of clause 44, wherein the at least one processor is further configured to: receive, via the at least one transceiver, from each UE in the subset of UEs an indication that the UE satisfies the one or more selection criteria.

[0231] Clause 46. A network entity according to clause 45, wherein, in response to receiving an indication that the UE satisfies the one or more selection criteria, the machine learning model is sent to the UE.

[0232] Clause 47. A network entity according to any one of clauses 44 to 46, wherein at least one processor is further configured to: before training the machine learning model, send a configuration to each UE in the UE set via at least one transceiver to report whether the UE satisfies the one or more selection criteria; or send a configuration to each UE in the UE set via at least one transceiver to train the machine learning model based on a determination that the UE satisfies the one or more selection criteria and not report whether the UE satisfies the one or more selection criteria.

[0233] Clause 48. A network entity according to any one of clauses 44 to 47, wherein at least one processor is further configured to: send, via at least one transceiver, to each UE in the set of UEs a configuration of a time period during which to monitor the values ​​of the one or more selection criteria to determine whether the UE satisfies the one or more selection criteria.

[0234] Clause 49. A network entity according to any one of clauses 44 to 48, wherein at least one processor is further configured to: send, via at least one transceiver, to each UE in the subset of UEs a configuration of a time period during which to train the machine learning model.

[0235] Clause 50. A network entity according to clause 49, wherein the machine learning model is trained based on training data collected by a subset of UEs during a time period.

[0236] Clause 51. The network entity of clause 50, wherein the at least one processor is further configured to: send, via the at least one transceiver, to each UE in the subset of UEs a configuration of a type of training data to be collected.

[0237] Clause 52. A network entity according to any one of clauses 49 to 51, wherein the time period comprises: one or more update iterations of the machine learning model, or a time window.

[0238] Clause 53. A network entity according to any one of clauses 44 to 52, wherein the updated parameters include updated weights of the machine learning model, updated gradients of the machine learning model, or both.

[0239] Clause 54. A network entity according to any one of clauses 44 to 53, wherein the one or more selection criteria include: an area identifier criterion, a coverage area criterion, a local data set size criterion, a training load balancing criterion, a UE training processing capability criterion, a communication channel condition criterion, a test set performance criterion, or any combination thereof.

[0240] Clause 55. A network entity according to any one of clauses 44 to 54, wherein the machine learning model is a radio frequency fingerprint (RFFP) based machine learning model.

[0241] Clause 56. The network entity of any of clauses 44 to 55, wherein the network entity is a location server, an edge server, or a model repository server.

[0242] Clause 57. A user equipment (UE), comprising: a component for receiving one or more selection criteria from a network entity for determining whether the UE is to participate in training a machine learning model; a component for determining whether the UE satisfies the one or more selection criteria during a first time period; and a component for sending updated parameters of the machine learning model to the network entity after a second time period, wherein the machine learning model is updated during the second time period based on a determination that the UE satisfies the one or more selection criteria.

[0243] Clause 58. The UE of clause 57, further comprising: means for sending an indication to a network entity that the UE satisfies the one or more selection criteria.

[0244] Clause 59. A UE according to any one of clauses 57 to 58, further comprising: a component for receiving a configuration from a network entity for reporting whether the UE satisfies the one or more selection criteria before the machine learning model is updated; or a component for receiving a configuration from a network entity for updating the machine learning model based on a determination that the UE satisfies the one or more selection criteria and not reporting whether the UE satisfies the one or more selection criteria.

[0245] Clause 60. A UE according to any of clauses 57 to 59, further comprising means for receiving a configuration of the first time period, the second time period, or both from a network entity.

[0246] Clause 61. A UE according to any one of clauses 57 to 60, further comprising: a component for receiving a machine learning model from a network entity.

[0247] Clause 62. A UE according to clause 61, wherein the machine learning model is received at the following time: before the first time period, or after the first time period and before the machine learning model is updated.

[0248] Clause 63. A UE according to any one of clauses 57 to 62, further comprising: a component for performing multiple repetitions of determining whether the UE satisfies the one or more selection criteria and updating the machine learning model.

[0249] Clause 64. The UE according to Clause 63 further includes: a component for receiving a new machine learning model from a network entity for each repetition in multiple repetitions; or a component for receiving a single machine learning model from a network entity for multiple repetitions, wherein the machine learning model is a single machine learning model.

[0250] Clause 65. A UE according to any one of clauses 57 to 64, wherein the machine learning model is trained based on training data collected by the UE during the second time period.

[0251] Clause 66. The UE of clause 65, further comprising: means for receiving configuration of the type of training data to be collected from a network entity.

[0252] Clause 67. A UE according to any one of clauses 57 to 66, wherein the second time period comprises: one or more update iterations of the machine learning model, or a time window.

[0253] Clause 68. A UE according to any one of clauses 57 to 67, wherein the updated parameters include updated weights of the machine learning model, updated gradients of the machine learning model, or both.

[0254] Clause 69. A UE according to any one of clauses 57 to 68, wherein the one or more selection criteria include: an area identifier criterion, a coverage area criterion, a local data set size criterion, a training load balancing criterion, a UE training processing capability criterion, a communication channel condition criterion, a test set performance criterion, or any combination thereof.

[0255] Clause 70. A UE according to any one of clauses 57 to 69, wherein the machine learning model is a radio frequency fingerprint (RFFP) based machine learning model.

[0256] Clause 71. The UE of any of clauses 57 to 70, wherein the network entity is a location server, an edge server, or a model repository server.

[0257] Clause 72. A network entity comprising: a component for sending one or more selection criteria for determining whether the set of UEs is to participate in training a machine learning model to a set of user equipment (UEs); a component for sending the machine learning model to at least one subset of UEs in the set of UEs that meet the one or more selection criteria; a component for receiving updated parameters of the machine learning model from each UE in the subset of UEs; and a component for updating the machine learning model based on the updated parameters received from each UE in the subset of UEs.

[0258] Clause 73. The network entity of clause 72, further comprising: means for receiving, from each UE in the subset of UEs, an indication that the UE satisfies the one or more selection criteria.

[0259] Clause 74. A network entity according to clause 73, wherein, in response to receiving an indication that the UE satisfies the one or more selection criteria, the machine learning model is sent to the UE.

[0260] Clause 75. A network entity according to any one of clauses 72 to 74 further includes: a component for sending a configuration for reporting whether the UE satisfies the one or more selection criteria to each UE in the UE set before training the machine learning model; or a component for sending a configuration for training the machine learning model to each UE in the UE set based on a determination that the UE satisfies the one or more selection criteria and not reporting whether the UE satisfies the one or more selection criteria.

[0261] Clause 76. A network entity according to any of clauses 72 to 75, further comprising: means for sending to each UE in the set of UEs a configuration of a time period during which to monitor the value of one or more selection criteria to determine whether the UE satisfies the one or more selection criteria.

[0262] Clause 77. A network entity according to any one of clauses 72 to 76, further comprising: a component for sending to each UE in the subset of UEs a configuration of a time period during which to train a machine learning model.

[0263] Clause 78. A network entity according to clause 77, wherein the machine learning model is trained based on training data collected by a subset of UEs during the time period.

[0264] Clause 79. The network entity of clause 78, further comprising: means for sending to each UE in the subset of UEs a configuration of the type of training data to be collected.

[0265] Clause 80. A network entity according to any one of clauses 77 to 79, wherein the time period comprises: one or more update iterations of the machine learning model, or a time window.

[0266] Clause 81. A network entity according to any one of clauses 72 to 80, wherein the updated parameters include updated weights of the machine learning model, updated gradients of the machine learning model, or both.

[0267] Clause 82. A network entity according to any one of clauses 72 to 81, wherein the one or more selection criteria include: an area identifier criterion, a coverage area criterion, a local data set size criterion, a training load balancing criterion, a UE training processing capability criterion, a communication channel condition criterion, a test set performance criterion, or any combination thereof.

[0268] Clause 83. A network entity according to any one of clauses 72 to 82, wherein the machine learning model is a radio frequency fingerprint (RFFP) based machine learning model.

[0269] Clause 84. The network entity of any of clauses 72 to 83, wherein the network entity is a location server, an edge server, or a model repository server.

[0270] Clause 85. A non-transitory computer-readable medium storing computer-executable instructions, which, when executed by a user equipment (UE), causes the UE to: receive one or more selection criteria from a network entity for determining whether the UE is to participate in training a machine learning model; determine whether the UE satisfies the one or more selection criteria during a first time period; and after a second time period, send updated parameters of the machine learning model to the network entity, wherein the machine learning model is updated during the second time period based on a determination that the UE satisfies the one or more selection criteria.

[0271] Clause 86. A non-transitory computer-readable medium according to any one of clauses 57 to 85, further comprising computer-executable instructions which, when executed by a UE, cause the UE to: send an indication to a network entity that the UE satisfies the one or more selection criteria.

[0272] Clause 87. A non-transitory computer-readable medium according to any one of clauses 57 to 86, further comprising computer-executable instructions, which, when executed by a UE, cause the UE to: receive a configuration from a network entity for reporting whether the UE satisfies the one or more selection criteria before the machine learning model is updated; or receive a configuration from a network entity for updating the machine learning model based on a determination that the UE satisfies the one or more selection criteria and not reporting whether the UE satisfies the one or more selection criteria.

[0273] Clause 88. The non-transitory computer-readable medium of any of clauses 57 to 87, further comprising computer-executable instructions which, when executed by a UE, cause the UE to: receive a configuration of the first time period, the second time period, or both from a network entity.

[0274] Clause 89. A non-transitory computer-readable medium according to any one of clauses 57 to 88, further comprising computer-executable instructions which, when executed by a UE, cause the UE to: receive a machine learning model from a network entity.

[0275] Clause 90. A non-transitory computer-readable medium according to any one of clauses 61 to 89, wherein the machine learning model is received at the following time: before the first time period, or after the first time period and before the machine learning model is updated.

[0276] Clause 91. A non-transitory computer-readable medium according to any one of clauses 57 to 90, further comprising computer-executable instructions, which, when executed by a UE, cause the UE to: perform multiple repetitions of determining whether the UE satisfies the one or more selection criteria and updating the machine learning model.

[0277] Clause 92. A non-transitory computer-readable medium according to any one of clauses 63 to 91, further comprising computer-executable instructions, which, when executed by a UE, cause the UE to: receive a new machine learning model from a network entity for each of multiple repetitions; or receive a single machine learning model from a network entity for multiple repetitions, wherein the machine learning model is a single machine learning model.

[0278] Clause 93. A non-transitory computer-readable medium according to any of clauses 57 to 92, wherein the machine learning model is trained based on training data collected by the UE during the second time period.

[0279] Clause 94. The non-transitory computer-readable medium of any of clauses 65 to 93, further comprising computer-executable instructions which, when executed by a UE, cause the UE to: receive, from a network entity, a configuration of a type of training data to be collected.

[0280] Clause 95. A non-transitory computer-readable medium according to any one of clauses 57 to 94, wherein the second time period comprises: one or more update iterations of the machine learning model, or a time window.

[0281] Clause 96. A non-transitory computer-readable medium according to any of clauses 57 to 95, wherein the updated parameters include updated weights of the machine learning model, updated gradients of the machine learning model, or both.

[0282] Clause 97. A non-transitory computer-readable medium according to any of clauses 57 to 96, wherein one or more selection criteria include: a region identifier criterion, a coverage area criterion, a local data set size criterion, a training load balancing criterion, a UE training processing capability criterion, a communication channel condition criterion, a test set performance criterion, or any combination thereof.

[0283] Clause 98. A non-transitory computer-readable medium according to any one of clauses 57 to 97, wherein the machine learning model is a radio frequency fingerprint (RFFP) based machine learning model.

[0284] Clause 99. The non-transitory computer-readable medium of any of clauses 57 to 98, wherein the network entity is a location server, an edge server, or a model repository server.

[0285] Clause 100. A non-transitory computer-readable medium storing computer-executable instructions which, when executed by a network entity, causes the network entity to: send one or more selection criteria to a set of user equipment (UEs) for determining whether the set of UEs is to participate in training a machine learning model; send the machine learning model to at least one subset of UEs in the set of UEs that meet the one or more selection criteria; receive updated parameters of the machine learning model from each UE in the subset of UEs; and update the machine learning model based on the updated parameters received from each UE in the subset of UEs.

[0286] Clause 101. A non-transitory computer-readable medium according to any one of clauses 72 to 100, further comprising computer-executable instructions which, when executed by a network entity, cause the network entity to: receive, from each UE in the subset of UEs, an indication that the UE satisfies the one or more selection criteria.

[0287] Clause 102. A non-transitory computer-readable medium according to any one of clauses 73 to 101, wherein, in response to receiving an indication that the UE satisfies the one or more selection criteria, the machine learning model is sent to the UE.

[0288] Clause 103. A non-transitory computer-readable medium according to any one of clauses 72 to 102, further comprising computer-executable instructions which, when executed by a network entity, cause the network entity to: before training the machine learning model, send a configuration to each UE in the UE set for reporting whether the UE satisfies the one or more selection criteria; or send a configuration to each UE in the UE set for training the machine learning model based on a determination that the UE satisfies the one or more selection criteria and not reporting whether the UE satisfies the one or more selection criteria.

[0289] Clause 104. A non-transitory computer-readable medium according to any one of clauses 72 to 103, further comprising computer-executable instructions which, when executed by a network entity, cause the network entity to: send to each UE in the set of UEs a configuration of a time period during which to monitor the value of the one or more selection criteria to determine whether the UE satisfies the one or more selection criteria.

[0290] Clause 105. A non-transitory computer-readable medium according to any one of clauses 72 to 104, further comprising computer-executable instructions which, when executed by a network entity, cause the network entity to: send to each UE in a subset of UEs a configuration of a time period during which to train a machine learning model.

[0291] Clause 106. A non-transitory computer-readable medium according to any of clauses 77 to 105, wherein the machine learning model is trained based on training data collected by the subset of UEs during the time period.

[0292] Clause 107. The non-transitory computer-readable medium of any of clauses 78 to 106, further comprising computer-executable instructions which, when executed by a network entity, cause the network entity to: send a configuration of a type of training data to be collected to each UE in the subset of UEs.

[0293] Clause 108. A non-transitory computer-readable medium according to any of clauses 77 to 107, wherein the time period comprises: one or more update iterations of the machine learning model, or a time window.

[0294] Clause 109. A non-transitory computer-readable medium according to any one of clauses 72 to 108, wherein the updated parameters include updated weights of the machine learning model, updated gradients of the machine learning model, or both.

[0295] Clause 110. A non-transitory computer-readable medium according to any of clauses 72 to 109, wherein the one or more selection criteria include: a region identifier criterion, a coverage area criterion, a local data set size criterion, a training load balancing criterion, a UE training processing capability criterion, a communication channel condition criterion, a test set performance criterion, or any combination thereof.

[0296] Clause 111. A non-transitory computer-readable medium according to any one of clauses 72 to 110, wherein the machine learning model is a radio frequency fingerprint (RFFP) based machine learning model.

[0297] Clause 112. The non-transitory computer-readable medium of any of Clauses 72 to 111, wherein the network entity is a location server, an edge server, or a model repository server.

[0298] Those skilled in the art will appreciate that information and signals may be represented using any of a variety of different techniques and technologies. For example, data, instructions, commands, information, signals, bits, symbols, and chips referenced throughout the above description may be represented by voltages, currents, electromagnetic waves, magnetic fields or particles, optical fields or particles, or any combination thereof.

[0299] Further, it will be appreciated by those skilled in the art that the various illustrative logic blocks, modules, circuits, and algorithmic steps described in conjunction with the aspects disclosed herein may be implemented as electronic hardware, computer software, or a combination of the two. In order to clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been generally described above according to their functions. Whether such functions are implemented as hardware or software depends on specific applications and design restrictions on the entire system. Technicians may implement the described functions in different ways for each specific application, but such implementation decisions should not be interpreted as resulting in departure from the scope of the present disclosure.

[0300] The various exemplary logical blocks, modules, and circuits described in conjunction with the various aspects disclosed herein may be implemented or executed with a general purpose processor, a digital signal processor (DSP), an ASIC, a field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general purpose processor may be a microprocessor, but alternatively, the processor may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, for example, a combination of a DSP and a microprocessor, a plurality of microprocessors, a combination of one or more microprocessors and a DSP core, or any other such configuration.

[0301] The methods, sequences and / or algorithms described in conjunction with the various aspects disclosed herein may be implemented directly in hardware, in a software module executed by a processor, or in a combination of the two. The software module may be located in a random access memory (RAM), a flash memory, a read-only memory (ROM), an erasable programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM), a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. An example storage medium is coupled to a processor so that the processor can read information from the storage medium and write information to the storage medium. In an alternative, the storage medium may be integrated into a processor. The processor and the storage medium may be located in an ASIC. The ASIC may be located in a user terminal (e.g., a UE). In an alternative, the processor and the storage medium may be located in a user terminal as discrete components.

[0302] In one or more example aspects, the functions described may be implemented with hardware, software, firmware, or any combination thereof. If implemented by software, these functions may be stored or sent as one or more instructions or codes in a computer-readable medium. Computer-readable media include computer storage media and communication media, and communication media include any media that facilitates the transfer of computer programs from one place to another. Storage media may be any available media that can be accessed by a computer. As an example and not limitation, such computer-readable media may include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program codes in the form of instructions or data structures and can be accessed by a computer. In addition, any connection is properly referred to as a computer-readable medium. For example, if the software is sent from a website, server, or other remote source using a coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technology (such as infrared, radio, and microwave), then coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technology (such as infrared, radio, and microwave) are all included in the definition of the medium. As used herein, disk and disc include compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk and Blu-ray disc, where disks usually reproduce data magnetically, while discs reproduce data optically with lasers. Combinations of the above should also be included within the scope of computer readable media.

[0303] Although the above disclosure shows the illustrative aspects of the present disclosure, it should be noted that various changes and modifications may be made to this document without departing from the scope of the present disclosure as defined by the appended claims. The functions, steps and / or actions of the method claims according to the aspects of the disclosure described herein do not need to be performed in any particular order. In addition, although the elements of the present disclosure may be described or stated in singular form, plural forms are also contemplated unless expressly stated to be limited to the singular form.

Claims

1. A method for training a machine learning model performed by a user equipment UE, comprising: receiving, from a network entity, one or more selection criteria for determining whether the UE is to participate in training the machine learning model; determining whether the UE satisfies the one or more selection criteria during a first time period; as well as After a second time period, updated parameters of the machine learning model are sent to the network entity, wherein the machine learning model is updated during the second time period based on a determination that the UE satisfies the one or more selection criteria.

2. The method according to claim 1, further comprising: An indication is sent to the network entity that the UE satisfies the one or more selection criteria.

3. The method according to claim 1, further comprising: receiving, before the machine learning model is updated, a configuration from the network entity for reporting whether the UE satisfies the one or more selection criteria; or Receiving from the network entity a configuration for updating the machine learning model based on a determination that the UE satisfies the one or more selection criteria and not reporting whether the UE satisfies the one or more selection criteria.

4. The method according to claim 1, further comprising: A configuration of the first time period, the second time period, or both is received from the network entity.

5. The method according to claim 1, further comprising: The machine learning model is received from the network entity.

6. The method according to claim 5, wherein: The machine learning model was received at: Before the first period of time, or After the first time period and before the machine learning model is updated.

7. The method according to claim 1, further comprising: Perform multiple iterations of determining whether the UE meets the one or more selection criteria and updating the machine learning model.

8. The method according to claim 7, further comprising: For each of the plurality of iterations, receiving a new machine learning model from the network entity; or For the multiple iterations, receiving a single machine learning model from the network entity, wherein the machine learning model is the single machine learning model.

9. The method according to claim 1, wherein: The machine learning model is trained based on training data collected by the UE during the second time period.

10. The method according to claim 9, further comprising: A configuration of the type of the training data to be collected is received from the network entity.

11. The method according to claim 1, wherein: The second time period includes: one or more update iterations of the machine learning model, or Time window.

12. The method according to claim 1, wherein: The updated parameters include updated weights of the machine learning model, updated gradients of the machine learning model, or both.

13. The method according to claim 1, wherein: The one or more selection criteria include: Region Identifier Standard, Coverage area standards, Local dataset size standard, Training load balancing standards, UE training processing capability standards, Communication channel condition standards, Test set performance criteria, or Any combination thereof.

14. The method according to claim 1, wherein: The machine learning model is a machine learning model based on radio frequency fingerprint RFFP.

15. The method according to claim 1, wherein: The network entity is a location server, an edge server, or a model repository server.

16. A method for training a machine learning model performed by a network entity, comprising: Sending one or more selection criteria to a set of user equipments UE for determining whether the set of UEs are to participate in training the machine learning model; sending the machine learning model to at least one subset of UEs in the set of UEs that satisfy the one or more selection criteria; receiving updated parameters of the machine learning model from each UE in the subset of UEs; as well as The machine learning model is updated based on the updated parameters received from each UE in the subset of UEs.

17. The method according to claim 16, further comprising: An indication is received from each UE in the subset of UEs that the UE satisfies the one or more selection criteria.

18. The method according to claim 17, wherein: In response to receiving the indication that the UE satisfies the one or more selection criteria, the machine learning model is sent to the UE.

19. The method according to claim 16, further comprising: Prior to training the machine learning model, sending to each UE in the set of UEs a configuration for reporting whether the UE satisfies the one or more selection criteria; or Sending to each UE in the set of UEs a configuration for training the machine learning model based on a determination that the UE satisfies the one or more selection criteria and not reporting whether the UE satisfies the one or more selection criteria.

20. The method of claim 16, further comprising: A configuration of a time period is sent to each UE in the set of UEs, during which values ​​of the one or more selection criteria are monitored to determine whether the UE satisfies the one or more selection criteria.

21. The method of claim 16, further comprising: A configuration of a time period is sent to each UE in the subset of UEs, during which the machine learning model is trained.

22. The method according to claim 21, wherein: The machine learning model is trained based on training data collected by the subset of UEs during the time period.

23. The method according to claim 22, further comprising: A configuration of the type of training data to be collected is sent to each UE in the subset of UEs.

24. The method according to claim 21, wherein: The time period includes: one or more update iterations of the machine learning model, or Time window.

25. The method of claim 16, wherein: The updated parameters include updated weights of the machine learning model, updated gradients of the machine learning model, or both.

26. The method of claim 16, wherein: The one or more selection criteria include: Region Identifier Standard, Coverage area standards, Local dataset size standard, Training load balancing standards, UE training processing capability standards, Communication channel condition standards, Test set performance criteria, or Any combination thereof.

27. The method of claim 16, wherein: The machine learning model is a machine learning model based on radio frequency fingerprint RFFP.

28. The method of claim 16, wherein: The network entity is a location server, an edge server, or a model repository server.

29. A user equipment UE, comprising: Memory; at least one transceiver; as well as at least one processor, communicatively coupled to the memory and the at least one transceiver, the at least one processor configured to: receiving, via the at least one transceiver, from a network entity one or more selection criteria for determining whether the UE is to participate in training a machine learning model; determining whether the UE satisfies the one or more selection criteria during a first time period; as well as After a second time period, updated parameters of the machine learning model are sent to the network entity via the at least one transceiver, wherein the machine learning model is updated during the second time period based on a determination that the UE satisfies the one or more selection criteria.

30. The UE according to claim 29, wherein: The at least one processor is further configured to: An indication that the UE satisfies the one or more selection criteria is sent to the network entity via the at least one transceiver.

31. The UE according to claim 29, wherein: The at least one processor is further configured to: receiving, before the machine learning model is updated, from the network entity via the at least one transceiver, a configuration for reporting whether the UE satisfies the one or more selection criteria; or Receiving, via the at least one transceiver, from the network entity a configuration for updating the machine learning model based on a determination that the UE satisfies the one or more selection criteria and not reporting whether the UE satisfies the one or more selection criteria.

32. The UE according to claim 29, wherein: The at least one processor is further configured to: A configuration of the first time period, the second time period, or both is received from the network entity via the at least one transceiver.

33. The UE according to claim 29, wherein: The at least one processor is further configured to: The machine learning model is received from the network entity via the at least one transceiver.

34. The UE according to claim 33, wherein: The machine learning model was received at: Before the first period of time, or After the first time period and before the machine learning model is updated.

35. The UE according to claim 29, wherein: The at least one processor is further configured to: Perform multiple iterations of determining whether the UE meets the one or more selection criteria and updating the machine learning model.

36. The UE according to claim 35, wherein: The at least one processor is further configured to: For each repetition of the plurality of repetitions, receiving a new machine learning model from the network entity via the at least one transceiver; or For the multiple repetitions, a single machine learning model is received from the network entity via the at least one transceiver, wherein the machine learning model is the single machine learning model.

37. The UE according to claim 29, wherein: The machine learning model is trained based on training data collected by the UE during the second time period.

38. The UE according to claim 37, wherein: The at least one processor is further configured to: A configuration of a type of the training data to be collected is received from a network entity via the at least one transceiver.

39. The UE according to claim 29, wherein: The second time period includes: one or more update iterations of the machine learning model, or Time window.

40. The UE according to claim 29, wherein: The updated parameters include updated weights of the machine learning model, updated gradients of the machine learning model, or both.

41. The UE according to claim 29, wherein: The one or more selection criteria include: Region Identifier Standard, Coverage area standards, Local dataset size standard, Training load balancing standards, UE training processing capability standards, Communication channel condition standards, Test set performance criteria, or Any combination thereof.

42. The UE according to claim 29, wherein: The machine learning model is a machine learning model based on radio frequency fingerprint RFFP.

43. The UE according to claim 29, wherein: The network entity is a location server, an edge server, or a model repository server.

44. A network entity comprising: Memory; at least one transceiver; as well as at least one processor, communicatively coupled to the memory and the at least one transceiver, the at least one processor configured to: Sending, via the at least one transceiver, to a set of user equipments (UEs), one or more selection criteria for determining whether the set of UEs are to participate in training the machine learning model; transmitting, via the at least one transceiver, the machine learning model to at least a subset of UEs in the set of UEs that satisfy the one or more selection criteria; receiving, via the at least one transceiver, updated parameters of the machine learning model from each UE in the subset of UEs; as well as The machine learning model is updated based on the updated parameters received from each UE in the subset of UEs.

45. The network entity according to claim 44, wherein: The at least one processor is further configured to: An indication is received from each UE in the subset of UEs via the at least one transceiver that the UE satisfies the one or more selection criteria.

46. ​​The network entity according to claim 45, wherein: In response to receiving the indication that the UE satisfies the one or more selection criteria, the machine learning model is sent to the UE.

47. The network entity according to claim 44, wherein: The at least one processor is further configured to: Prior to training the machine learning model, sending, via the at least one transceiver, to each UE in the set of UEs a configuration for reporting whether the UE satisfies the one or more selection criteria; or A configuration is sent, via the at least one transceiver, to each UE in the set of UEs for training the machine learning model based on a determination that the UE satisfies the one or more selection criteria and not reporting whether the UE satisfies the one or more selection criteria.

48. The network entity of claim 44, wherein: The at least one processor is further configured to: A configuration of a time period is transmitted to each UE in the set of UEs via the at least one transceiver, during which the value of the one or more selection criteria is monitored to determine whether the UE satisfies the one or more selection criteria.

49. The network entity of claim 44, wherein: The at least one processor is further configured to: A configuration of a time period is sent to each UE in the subset of UEs via the at least one transceiver, during which the machine learning model is trained.

50. The network entity according to claim 49, wherein: The machine learning model is trained based on training data collected by the subset of UEs during the time period.

51. The network entity according to claim 50, wherein: The at least one processor is further configured to: A configuration of the type of training data to be collected is sent to each UE in the subset of UEs via the at least one transceiver.

52. The network entity of claim 49, wherein: The time period includes: one or more update iterations of the machine learning model, or Time window.

53. The network entity of claim 44, wherein: The updated parameters include updated weights of the machine learning model, updated gradients of the machine learning model, or both.

54. The network entity of claim 44, wherein: The one or more selection criteria include: Region Identifier Standard, Coverage area standards, Local dataset size standard, Training load balancing standards, UE training processing capability standards, Communication channel condition standards, Test set performance criteria, or Any combination thereof.

55. The network entity of claim 44, wherein: The machine learning model is a machine learning model based on radio frequency fingerprint RFFP.

56. The network entity of claim 44, wherein: The network entity is a location server, an edge server, or a model repository server.

57. A user equipment UE, comprising: means for receiving from a network entity one or more selection criteria for determining whether the UE is to participate in training a machine learning model; means for determining whether the UE satisfies the one or more selection criteria during a first time period; as well as and means for sending updated parameters of the machine learning model to the network entity after a second time period, wherein the machine learning model is updated during the second time period based on a determination that the UE satisfies the one or more selection criteria.

58. A network entity comprising: A component for sending one or more selection criteria to a set of user equipments (UEs) for determining whether the set of UEs are to participate in training a machine learning model; means for sending the machine learning model to at least a subset of UEs in the set of UEs that satisfy the one or more selection criteria; means for receiving updated parameters of the machine learning model from each UE in the subset of UEs; as well as A component for updating the machine learning model based on the updated parameters received from each UE in the subset of UEs.

59. A non-transitory computer-readable medium storing computer-executable instructions, which, when executed by a user equipment (UE), causes the UE to: receiving, from a network entity, one or more selection criteria for determining whether the UE is to participate in training the machine learning model; determining whether the UE satisfies the one or more selection criteria during a first time period; as well as After a second time period, updated parameters of the machine learning model are sent to the network entity, wherein the machine learning model is updated during the second time period based on a determination that the UE satisfies the one or more selection criteria.

60. A non-transitory computer-readable medium storing computer-executable instructions that, when executed by a network entity, cause the network entity to: Sending one or more selection criteria to a set of user equipments (UEs) for determining whether the set of UEs are to participate in training a machine learning model; sending the machine learning model to at least one subset of UEs in the set of UEs that satisfy the one or more selection criteria; receiving updated parameters of the machine learning model from each UE in the subset of UEs; as well as The machine learning model is updated based on the updated parameters received from each UE in the subset of UEs.

Images