Machine learning model management and assistance information

Abstract

Certain aspects of the present disclosure provide a method of wireless communication by a user equipment (UE), the method generally including obtaining, from a network entity, a performance report for a machine learning (ML) model running on at least one of the UE or the network entity; and participating in a change to the ML model based on the performance report.

Description

Technical Field

[0001] Aspects of the present disclosure relate to wireless communications, and more particularly to techniques for enhancing machine learning (ML) based models used in wireless communication systems. Background Art

[0002] Wireless communication systems are widely deployed to provide various telecommunication services such as telephony, video, data, messaging, broadcasts or other similar types of services. These wireless communication systems may employ multiple-access technologies capable of supporting communication with multiple users by sharing available wireless communication system resources with the multiple users.

[0003] Although wireless communication systems have made tremendous technical progress over the years, challenges still exist. For example, complex and dynamic environments can still attenuate or block signals between wireless transmitters and wireless receivers. Therefore, there is a continuous desire to improve the technical performance of wireless communication systems, including, for example: improving the speed and data carrying capacity of communications, improving the efficiency of using shared communication media, reducing the power used by transmitters and receivers when performing communications, improving the reliability of wireless communications, avoiding redundant transmission and / or reception and related processing, improving the coverage area of ​​wireless communications, increasing the number and types of devices that can access wireless communication systems, increasing the ability of different types of devices to communicate with each other, increasing the number and types of wireless communication media available for use, etc. Therefore, there is a need to further improve wireless communication systems to overcome the above-mentioned technical challenges and other challenges. Summary of the invention

[0004] One aspect provides a method of wireless communication (e.g., by a user equipment (UE)). The method includes: obtaining a performance report for a machine learning (ML) model running on at least one of the UE or the network entity from a network entity; and engaging in changes to the ML model based on the performance report.

[0005] Another aspect provides a method of wireless communication by a UE. The method includes: generating a performance report for an ML model running on at least one of the UE or a network entity; and engaging in changes to the ML model based on the performance report.

[0006] Another aspect provides a method of wireless communication by a network entity. The method includes: sending a performance report for an ML model running on at least one of a UE or the network entity; and engaging in changes to the ML model based on the performance report.

[0007] Another aspect provides a method of wireless communication by a network entity. The method includes: receiving a performance report for an ML model running on at least one of a UE or the network entity; and engaging in changes to the ML model based on the performance report.

[0008] Other aspects provide: an apparatus operable to, configured to, or otherwise adapted to perform any one or more of the foregoing methods and / or those described elsewhere herein; a non-transitory computer-readable medium comprising instructions that, when executed by a processor of the apparatus, cause the apparatus to perform the foregoing methods and those described elsewhere herein; a computer program product embodied on a computer-readable storage medium comprising code for performing the foregoing methods and those described elsewhere herein; and / or an apparatus comprising components for performing the foregoing methods and those described elsewhere herein. By way of example, an apparatus may include a processing system, a device having a processing system, or a processing system cooperating through one or more networks.

[0009] For purposes of illustration, the following description and drawings set forth certain features. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The accompanying drawings depict certain features of the various aspects described herein and should not be considered to limit the scope of the disclosure.

[0011] Figure 1 An example wireless communication network is depicted.

[0012] Figure 2 An example disaggregated base station architecture is depicted.

[0013] Figure 3 Aspects of an example base station and example user equipment are depicted.

[0014] Figure 4A , Figure 4B , Figure 4C and Figure 4D Various example aspects of data structures for a wireless communication network are described.

[0015] Figure 5 An example beam refinement procedure in accordance with certain aspects of the present disclosure is illustrated.

[0016] Figure 6 is a diagram illustrating example operations that may be performed for beam management.

[0017] Figure 7 The general functional framework applied to AI-enabled RAN intelligence is illustrated.

[0018] Figure 8 Depicts an example of monitoring the performance of a machine learning model over time.

[0019] Fig. 9 An example of an ML-based CSI feedback mechanism is depicted.

[0020] Fig.10 Describes the use Fig. 9 Examples of encoder input and decoder output for the ML-based CSI feedback mechanism.

[0021] Fig.11 Depicted are examples of network-side ML model retraining or switching decisions in accordance with aspects of the present disclosure.

[0022] Fig.12 Depicted are examples of UE-side ML model retraining or handover decisions in accordance with aspects of the present disclosure.

[0023] Fig.13 Examples of UE-side ML model performance monitoring according to aspects of the present disclosure are depicted.

[0024] Fig.14 A method for wireless communication is described.

[0025] Fig.15 A method for wireless communication is described.

[0026] Fig.16 A method for wireless communication is described.

[0027] Fig.17 A method for wireless communication is described.

[0028] Fig.18 Aspects of an example communications device are depicted. DETAILED DESCRIPTION

[0029] Aspects of the present disclosure provide apparatus, methods, processing systems, and computer-readable media for managing models for channel state estimation and feedback.

[0030] Understanding the channel state between devices communicating in a wireless communication system is an important aspect of improving wireless communication performance. Traditionally, many techniques have been used to measure channel state and report feedback so that performance can be improved. However, such conventional techniques are usually relatively slow, power-hungry, and static in approach.

[0031] Machine learning represents an opportunity to improve upon many conventional techniques for measuring channel state and reporting feedback. For example, machine learning models can reduce the number of resource elements required to estimate channel state and improve estimates of values ​​used when reporting channel state. However, because wireless environments tend to be extremely dynamic, it is important to be able to monitor the performance of machine learning models that implement key channel state measurement and feedback procedures and take remedial action if, for example, the model begins to perform poorly. Such remedial action may include, for example, falling back to a baseline model (e.g., a non-machine learning model) to perform various aspects until the machine learning model can be reconfigured to maintain optimal performance.

[0032] Aspects described herein as a whole relate to methods for monitoring performance and detecting failures associated with machine learning models for channel state measurement and feedback procedures. In various aspects, the network may be configured and / or the user equipment may implement various modes for monitoring model performance. In some aspects, model performance is monitored by determining output variance events and reporting such variance events to the network and / or using such variance events to determine when the model becomes unreliable or "failed". In some aspects, the model variance event may be an out-of-distribution (OOD) event, which generally refers to a machine learning model generating an uncertain output based on an input different from its training data.

[0033] By actively monitoring the performance of the machine learning model with or without network assistance and detecting failures with respect to the machine learning model, the various aspects described herein achieve the benefits of the machine learning model, such as faster, more power efficient, and more accurate operation, while mitigating the likelihood that the performance of the machine learning model will degrade over time. Such degradation may, for example, be caused by exposure of the machine learning model to new environments and new conditions that were not initially considered during the training of the machine learning model. In the context of the various examples described herein, this may include a user equipment using a machine learning model to perform channel estimation and predict channel state information feedback in a radio environment that is different from the environment considered during model training. Detecting such degradation allows reconfiguration (e.g., retraining) of the machine learning model to maintain state-of-the-art performance, while allowing fallback to a baseline model.

[0034] Thus, the aspects described herein that implement the robust use of machine learning models for channel state measurement and feedback procedures generally enhance wireless communication performance, and more specifically enhance wireless communication performance by reducing power usage, increasing battery life, improving spectral efficiency, reducing latency, and reducing network overhead, to name just a few technical improvements.

[0035] Introduction to wireless communication networks

[0036] The techniques and methods described herein can be used in various wireless communication networks. Although various aspects may be described herein using terms commonly associated with 3G, 4G and / or 5G wireless technologies, various aspects of the present disclosure may also be applicable to other communication systems and standards not explicitly mentioned herein.

[0037] Figure 1 An example of a wireless communication network 100 is depicted in which various aspects described herein may be implemented.

[0038] In general, the wireless communication network 100 includes various network entities (alternatively, network elements or network nodes). A network entity is typically a communication device and / or a communication function performed by a communication device (e.g., user equipment (UE), base station (BS), component of a BS, server, etc.). For example, various functions of a network and various devices associated with and interacting with the network may be considered network entities. In addition, the wireless communication network 100 includes ground aspects, such as ground-based network entities (e.g., BS 102), and non-ground aspects, such as satellites 140 and aircraft 145, which may include airborne network entities (e.g., one or more BSs) capable of communicating with other network elements (e.g., ground BSs) and user equipment.

[0039] In the depicted example, the wireless communication network 100 includes a BS 102, a UE 104, and one or more core networks (such as an evolved packet core (EPC) 160 and a 5G core (5GC) network 190) that interoperate to provide communication services over various communication links (including wired and wireless links).

[0040] Figure 1 Various example UEs 104 are depicted, which may more generally include: a cellular phone, a smart phone, a Session Initiation Protocol (SIP) phone, a laptop, a personal digital assistant (PDA), a satellite radio, a global positioning system, a multimedia device, a video device, a digital audio player, a camera, a game console, a tablet computer, a smart device, a wearable device, a vehicle, an electric meter, a gas pump, a large or small kitchen appliance, a healthcare device, an implant, a sensor / actuator, a display, an Internet of Things (IoT) device, an Always-On (AON) device, an edge processing device, or other similar devices. UE 104 may also be more generally referred to as a mobile device, a wireless device, a wireless communication device, a station, a mobile station, a subscriber station, a mobile subscriber station, a mobile unit, a subscriber unit, a wireless unit, a remote unit, a remote device, an access terminal, a mobile terminal, a wireless terminal, a remote terminal, a handset, and others.

[0041] BS 102 wirelessly communicates with (e.g., transmits signals to or receives signals from) UE 104 via communication link 120. Communication link 120 between BS 102 and UE 104 may include uplink (UL) (also referred to as a reverse link) transmissions from UE 104 to BS 102 and / or downlink (DL) (also referred to as a forward link) transmissions from BS 102 to UE 104. In various aspects, communication link 120 may use multiple-input multiple-output (MIMO) antenna technology, including spatial multiplexing, beamforming, and / or transmit diversity.

[0042] BS 102 may generally include: NodeB, enhanced NodeB (eNB), next generation enhanced NodeB (ng-eNB), next generation NodeB (gNB or gNodeB), access point, base transceiver station, radio base station, radio transceiver, transceiver function, transmission and reception point, and / or other. Each of BS 102 may provide communication coverage for a corresponding geographic coverage area 110, which may sometimes be referred to as a cell, and may overlap in some cases (e.g., a small cell 102' may have a coverage area 110' that overlaps with the coverage area 110 of a macro cell). For example, a BS may provide communication coverage for a macro cell (covering a relatively large geographic area), a pico cell (covering a relatively small geographic area, such as a stadium), a femto cell (relatively small geographic area (e.g., a home)), and / or other types of cells.

[0043] Although BS102 is depicted as a single communication device in various aspects, BS102 can be implemented in various configurations. For example, one or more components of the base station may be decomposed, including a central unit (CU), one or more distributed units (DU), one or more radio units (RU), a near real-time (near RT) RAN intelligent controller (RIC) or a non-real-time (non-RT) RIC, to name a few examples. In another example, various aspects of the base station may be virtualized. More generally, a base station (e.g., BS102) may include components located at a single physical location or components located at various physical locations. In an example in which the base station includes components located at various physical locations, the various components may each perform a function so that the various components together implement functionality similar to that of a base station located at a single physical location. In some aspects, a base station including components located at various physical locations may be referred to as a decomposed radio access network architecture (such as an open RAN (O-RAN) or virtualized RAN (VRAN) architecture). Figure 2 An example decomposed base station architecture is depicted and described.

[0044] Different BSs 102 within the wireless communication network 100 may also be configured to support different radio access technologies (such as 3G, 4G, and / or 5G). For example, a BS 102 configured for 4G LTE (collectively referred to as the Evolved Universal Mobile Telecommunications System (UMTS) Terrestrial Radio Access Network (E-UTRAN)) may interface with the EPC 160 via a first backhaul link 132 (e.g., an S1 interface). A BS 102 configured for 5G (e.g., 5G NR or Next Generation RAN (NG-RAN)) may interface with the 5GC 190 via a second backhaul link 184. The BSs 102 may communicate with each other directly or indirectly (e.g., via the EPC 160 or the 5GC 190) over a third backhaul link 134 (e.g., an X2 interface), which may be wired or wireless.

[0045] The wireless communication network 100 may subdivide the electromagnetic spectrum into various categories, frequency bands, channels, or other characteristics. In some aspects, the subdivision is provided based on wavelength and frequency, where frequency may also be referred to as a carrier, subcarrier, channel, tone, or subband. For example, 3GPP currently defines frequency range 1 (FR1) as including 410 MHz to 7125 MHz, which is often (interchangeably) referred to as "below 6 GHz". Similarly, 3GPP currently defines frequency range 2 (FR2) as including 24,250 MHz-52,600 MHz, which is sometimes (interchangeably) referred to as "millimeter wave" ("mmW" or "mmWave"). A base station (e.g., a mmWave base station such as BS180) configured to communicate using mmWave / near mmWave radio bands may utilize beamforming (e.g., 182) with a UE (e.g., 104) to improve path loss and range.

[0046] The communication link 120 between the BS 102 and, for example, the UE 104 may be through one or more carriers, which may have different bandwidths (e.g., 5 MHz, 10 MHz, 15 MHz, 20 MHz, 100 MHz, 400 MHz, and / or other MHz) and may be aggregated in various aspects. The carriers may or may not be adjacent to each other. The allocation of carriers may be asymmetric for DL ​​and UL (e.g., more or fewer carriers may be allocated for DL ​​than for UL).

[0047] Communications using higher frequency bands may have higher path loss and shorter range than communications at lower frequencies. Therefore, some base stations (e.g. Figure 1180) can utilize beamforming 182 with UE 104 to improve path loss and distance. For example, BS 180 and UE 104 can each include multiple antennas (such as antenna elements, antenna panels and / or antenna arrays) to facilitate beamforming. In some cases, BS 180 can send beamformed signals to UE 104 in one or more transmit directions 182'. UE 104 can receive beamformed signals from BS 180 in one or more receive directions 182". UE 104 can also send beamformed signals to BS 180 in one or more transmit directions 182". BS 180 can also receive beamformed signals from UE 104 in one or more receive directions 182'. BS 180 and UE 104 can then perform beam training to determine the best receive and transmit directions for each of BS 180 and UE 104. It is worth noting that the transmit direction and receive direction of BS 180 can be the same or different. Similarly, the transmit direction and receive direction of UE 104 may be the same or different.

[0048] Wireless communication network 100 also includes Wi-Fi AP 150 that communicates with Wi-Fi station (STA) 152 via communication link 154 in, for example, the 2.4 GHz and / or 5 GHz unlicensed spectrum.

[0049] Certain UEs 104 may communicate with each other using a device-to-device (D2D) communication link 158. The D2D communication link 158 may use one or more sidelink channels, such as a physical sidelink broadcast channel (PSBCH), a physical sidelink discovery channel (PSDCH), a physical sidelink shared channel (PSSCH), a physical sidelink control channel (PSCCH), and / or a physical sidelink feedback channel (PSFCH).

[0050] The EPC 160 may include various functional components, including: a mobility management entity (MME) 162, other MMEs 164, a serving gateway 166, a multimedia broadcast multicast service (MBMS) gateway 168, a broadcast multicast service center (BM-SC) 170, and / or a packet data network (PDN) gateway 172, such as in the depicted example. The MME 162 may communicate with a home subscriber server (HSS) 174. The MME 162 is a control node that handles signaling between the UE 104 and the EPC 160. Typically, the MME 162 provides bearer and connection management.

[0051] Typically, user Internet Protocol (IP) packets are transported through the Serving Gateway 166, which itself is connected to the PDN Gateway 172. The PDN Gateway 172 provides UE IP address allocation and other functions. The PDN Gateway 172 and the BM-SC 170 are connected to IP services 176, which may include, for example, the Internet, an intranet, an IP Multimedia Subsystem (IMS), a packet switched (PS) streaming media service, and / or other IP services.

[0052] The BM-SC 170 may provide functionality for MBMS user service configuration and delivery. The BM-SC 170 may serve as an entry point for content provider MBMS delivery, may be used to authorize and initiate MBMS bearer services within a public land mobile network (PLMN), and / or may be used to schedule MBMS delivery. The MBMS Gateway 168 may be used to distribute MBMS services to BSs 102 belonging to a multicast broadcast single frequency network (MBSFN) area broadcasting a specific service, and / or may be responsible for session management (start / stop) and for collecting eMBMS related charging information.

[0053] 5GC 190 may include various functional components, including: access and mobility management function (AMF) 192, other AMFs 193, session management function (SMF) 194, and user plane function (UPF) 195. AMF 192 may communicate with unified data management (UDM) 196.

[0054] AMF 192 is a control node that handles signaling between UE 104 and 5GC 190. AMF 192 provides, for example, Quality of Service (QoS) flow and session management.

[0055] Internet Protocol (IP) packets are delivered through UPF 195, which is connected to IP Services 197 and provides IP address allocation for UEs and other functions for 5GC 190. IP Services 197 may include, for example, the Internet, Intranet, IMS, PS streaming services, and / or other IP services.

[0056] In various aspects, a network entity or network node may be implemented as a converged base station, a decomposed base station, a component of a base station, an integrated access and backhaul (IAB) node, a relay node, a sidelink node, to name a few examples.

[0057] Figure 2An example disaggregated base station 200 architecture is depicted. The disaggregated base station 200 architecture may include one or more central units (CUs) 210 that may communicate directly with a core network 220 via a backhaul link, or indirectly with the core network 220 through one or more disaggregated base station units, such as a near real-time (near-RT) RAN intelligent controller (RIC) 225 via an E2 link, or a non-real-time (non-RT) RIC 215 associated with a service management and orchestration (SMO) framework 205, or both. The CU 210 may communicate with one or more distributed units (DUs) 230 via respective midhaul links, such as an F1 interface. The DU 230 may communicate with one or more radio units (RUs) 240 via respective fronthaul links. The RU 240 may communicate with respective UEs 104 via one or more radio frequency (RF) access links. In some implementations, a UE 104 may be served simultaneously by multiple RUs 240.

[0058] Each of the units (e.g., CU 210, DU 230, RU 240, and near-RT RIC 225, non-RT RIC 215, and SMO framework 205) 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 transmission medium or a wireless transmission medium. Each of the units 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 to one or more of the other units via a wired transmission medium. Additionally or alternatively, the unit 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 receive and send signals to one or more of the other units on a wireless transmission medium.

[0059] In some aspects, CU 210 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 as having an interface configured to communicate signals with other control functions hosted by CU 210. CU210 may be configured to handle user plane functionality (e.g., central unit-user plane (CU-UP)), control plane functionality (e.g., central unit-control plane (CU-CP)), or a combination thereof. In some specific implementations, CU 210 may be logically split into one or more CU-UP units and one or more CU-CP units. When implemented in an O-RAN configuration, the CU-UP unit may communicate bidirectionally with the CU-CP unit via an interface (such as an E1 interface). If necessary, CU 210 may be implemented to communicate with DU 230 for network control and signaling.

[0060] DU 230 may correspond to a logical unit that includes one or more base station functions for controlling the operation of one or more RUs 240. In some aspects, DU 230 may host one or more of a radio link control (RLC) layer, a medium 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.) depending at least in part on functional splits such as those defined by the 3rd Generation Partnership Project (3GPP). In some aspects, DU 230 may further host one or more low PHY layers. Each layer (or module) may be implemented with an interface that is configured to communicate signals with other layers (and modules) hosted by DU 230 or with control functions hosted by CU 210.

[0061] The lower layer functionality may be implemented by one or more RUs 240. In some deployments, a RU 240 controlled by a DU 230 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 a functional split (such as a lower layer functional split). In such an architecture, the RU 240 may be implemented to handle over-the-air (OTA) communications with one or more UEs 104. In some implementations, real-time and non-real-time aspects of control plane and user plane communications with the RU 240 may be controlled by the corresponding DU 230. In some scenarios, this configuration may enable the implementation of the DU 230 and the CU 210 in a cloud-based RAN architecture (such as a vRAN architecture).

[0062] The SMO framework 205 may be configured to support RAN deployment and provisioning of non-virtualized network elements and virtualized network elements. For non-virtualized network elements, the SMO framework 205 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 205 may be configured to interact with a cloud computing platform (such as an open cloud (O-cloud) 290) 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 210, DU 230, RU 240, and near-RT RIC 225. In some specific implementations, the SMO framework 205 may communicate with hardware aspects of the 4G RAN (such as an open eNB (O-eNB) 211) via the O1 interface. In addition, in some specific implementations, the SMO framework 205 may communicate directly with one or more RUs 240 via the O1 interface. The SMO framework 205 may also include a non-RT RIC 215 configured to support the functionality of the SMO framework 205 .

[0063] The non-RT RIC 215 may be configured to include logic functions that enable 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 225. The non-RT RIC 215 may be coupled to or in communication with the near-RT RIC 225 (such as via an A1 interface). The near-RT RIC 225 may be configured to include logic functions that enable near-real-time control and optimization of RAN elements and resources via data collection and actions through an interface (such as via an E2 interface) that connects one or more CUs 210, one or more DUs 230, or both, and the O-eNB with the near-RT RIC 225.

[0064] In some implementations, in order to generate an AI / ML model to be deployed in the near-RT RIC 225, the non-RT RIC 215 may receive parameters or external enrichment information from an external server. Such information may be utilized by the near-RT RIC 225 and may be received from a non-network data source or from a network function at the SMO framework 205 or the non-RT RIC 215. In some examples, the non-RT RIC 215 or the near-RT RIC 225 may be configured to tune RAN behavior or performance. For example, the non-RT RIC 215 may monitor long-term trends and patterns of performance and employ AI / ML models to perform corrective actions through the SMO framework 205 (such as via reconfiguration of O1) or via creation of RAN management policies (such as A1 policies).

[0065] Figure 3 Aspects of an example BS 102 and UE 104 are depicted.

[0066] In general, BS 102 includes various processors (e.g., 320, 330, 338, and 340), antennas 334a-t (collectively 334), transceivers 332a-t (collectively 332) including modulators and demodulators, and other aspects that enable wireless transmission of data (e.g., data source 312) and wireless reception of data (e.g., data sink 339). For example, BS 102 can transmit and receive data between BS 102 and UE 104. BS 102 includes controller / processor 340 that can be configured to implement various functions described herein related to wireless communication.

[0067] In general, the UE 104 includes various processors (e.g., 358, 364, 366, and 380), antennas 352a-r (collectively 352), transceivers 354a-r (collectively 354) including modulators and demodulators, and other aspects that enable wireless transmission of data (e.g., retrieved from a data source 362) and wireless reception of data (e.g., provided to a data sink 360). The UE 104 includes a controller / processor 380 that can be configured to implement various functions described herein related to wireless communications.

[0068] Regarding example downlink transmissions, BS 102 includes a transmit processor 320 that can receive data from a data source 312 and control information from a controller / processor 340. The control information can be for a physical broadcast channel (PBCH), a physical control format indicator channel (PCFICH), a physical HARQ indicator channel (PHICH), a physical downlink control channel (PDCCH), a group common PDCCH (GC PDCCH), and / or others. In some examples, the data can be for a physical downlink shared channel (PDSCH).

[0069] The transmit processor 320 may process (e.g., encode and symbol map) data and control information to obtain data symbols and control symbols, respectively. The transmit processor 320 may also generate reference symbols, such as for a primary synchronization signal (PSS), a secondary synchronization signal (SSS), a PBCH demodulation reference signal (DMRS), and a channel state information reference signal (CSI-RS).

[0070] The transmit (TX) multiple-input multiple-output (MIMO) processor 330 may perform spatial processing (e.g., pre-coding) on ​​data symbols, control symbols, and / or reference symbols, where applicable, and may provide an output symbol stream to a modulator (MOD) in a transceiver 332a-332t. Each modulator in the transceiver 332a-332t may process a corresponding output symbol stream to obtain an output sample stream. Each modulator may further process (e.g., convert to analog, amplify, filter, and up-convert) the output sample stream to obtain a downlink signal. The downlink signals from the modulators in the transceivers 332a-332t may be transmitted via antennas 334a-334t, respectively.

[0071] To receive downlink transmissions, UE 104 includes antennas 352a-352r that can receive downlink signals from BS 102 and can provide received signals to demodulators (DEMODs) in transceivers 354a-354r, respectively. Each demodulator in transceivers 354a-354r can condition (e.g., filter, amplify, downconvert, and digitize) a corresponding received signal to obtain input samples. Each demodulator can further process the input samples to obtain received symbols.

[0072] A MIMO detector 356 may obtain received symbols from all demodulators in transceivers 354a-354r, perform MIMO detection on the received symbols if applicable, and provide detected symbols. A receive processor 358 may process (e.g., demodulate, deinterleave, and decode) the detected symbols, provide decoded data for UE 104 to a data sink 360, and provide decoded control information to a controller / processor 380.

[0073] With respect to example uplink transmissions, the UE 104 also includes a transmit processor 364 that may receive and process data from a data source 362 (e.g., for a PUSCH) and control information from a controller / processor 380 (e.g., for a physical uplink control channel (PUCCH)). The transmit processor 364 may also generate reference symbols for reference signals (e.g., for a sounding reference signal (SRS)). The symbols from the transmit processor 364 may be pre-decoded by a TX MIMO processor 366, if applicable, further processed by a modulator in the transceivers 354a-354r (e.g., for SC-FDM), and transmitted to the BS 102.

[0074] At BS 102, uplink signals from UE 104 may be received by antennas 334a-t, processed by demodulators in transceivers 332a-332t, detected by MIMO detector 336 if applicable, and further processed by receive processor 338 to obtain decoded data and control information transmitted by UE 104. Receive processor 338 may provide decoded data to data sink 339 and decoded control information to controller / processor 340.

[0075] Memories 342 and 382 may store data and program codes for BS 102 and UE 104, respectively.

[0076] Scheduler 344 may schedule UEs for data transmission on the downlink and / or uplink.

[0077] In various aspects, the BS 102 may be described as sending and receiving various types of data associated with the methods described herein. In these contexts, "sending" may refer to various mechanisms for outputting data, such as outputting data from a data source 312, a scheduler 344, a memory 342, a transmit processor 320, a controller / processor 340, a TX MIMO processor 330, a transceiver 332a-t, an antenna 334a-t, and / or other aspects described herein. Similarly, "receiving" may refer to various mechanisms for obtaining data, such as obtaining data from an antenna 334a-t, a transceiver 332a-t, an RX MIMO detector 336, a controller / processor 340, a receive processor 338, a scheduler 344, a memory 342, and / or other aspects described herein.

[0078] In various aspects, the UE 104 may also be described as transmitting and receiving various types of data associated with the methods described herein. In these contexts, "transmitting" may refer to various mechanisms for outputting data, such as outputting data from a data source 362, a memory 382, ​​a transmit processor 364, a controller / processor 380, a TX MIMO processor 366, a transceiver 354a-t, an antenna 352a-t, and / or other aspects described herein. Similarly, "receiving" may refer to various mechanisms for obtaining data, such as obtaining data from an antenna 352a-t, a transceiver 354a-t, a RX MIMO detector 356, a controller / processor 380, a receive processor 358, a memory 382, ​​and / or other aspects described herein.

[0079] In some aspects, the processor may be configured to perform various operations (such as those associated with the methods described herein) and send (output) data to or receive (obtain) data from another interface configured to send or receive data, respectively.

[0080] Figure 4A , Figure 4B , Figure 4C and Figure 4D Describes a method for use in a wireless communication network such as Figure 1 Various aspects of the data structure of the wireless communication network 100).

[0081] Specifically, Figure 4A is a diagram 400 illustrating an example of a first subframe within a 5G (e.g., 5G NR) frame structure, Figure 4B is a diagram 430 illustrating an example of a DL channel within a 5G subframe, Figure 4C is a diagram 450 illustrating an example of a second subframe within a 5G frame structure, and Figure 4D FIG480 is a diagram illustrating an example of UL channels within a 5G subframe.

[0082] Wireless communication systems may utilize orthogonal frequency division multiplexing (OFDM) with a cyclic prefix (CP) on both the uplink and downlink. Such systems may also support half-duplex operation using time division duplex (TDD). OFDM and single carrier frequency division multiplexing (SC-FDM) can be used to maximize the system bandwidth (e.g., Figure 4B and Figure 4D The modulation symbols may be transmitted in the frequency domain using OFDM and / or in the time domain using SC-FDM.

[0083] The wireless communication frame structure may be frequency division duplex (FDD), where for a particular set of subcarriers, subframes within the set of subcarriers are dedicated to either DL or UL. The wireless communication frame structure may also be time division duplex (TDD), where for a particular set of subcarriers, subframes within the set of subcarriers are dedicated to both DL and UL.

[0084] exist Figure 4A and Figure 4C In the wireless communication frame structure, TDD is used, where D is DL, U is UL, and X is used flexibly between DL / UL. The UE can be configured with a time slot format (dynamically configured by DL control information (DCI) or semi-statically / statically configured by radio resource control (RRC) signaling) through the received time slot format indicator (SFI). In the depicted example, the 10ms frame is divided into 10 equally sized 1ms subframes. Each subframe may include one or more time slots. In some examples, each time slot may include 7 or 14 symbols, depending on the time slot format. The subframe may also include micro-time slots, which typically have fewer symbols than the entire time slot. Other wireless communication technologies may have different frame structures and / or different channels.

[0085] In certain aspects, the number of slots within a subframe is based on the slot configuration and parameter set. For example, for slot configuration 0, different parameter sets (μ) 0 to 5 allow for 1, 2, 4, 8, 16, and 32 slots per subframe, respectively. For slot configuration 1, different parameter sets 0 to 2 allow for 2, 4, and 8 slots per subframe, respectively. Thus, for slot configuration 0 and parameter set μ, there are 14 symbols / slot and 2μ slots / subframe. The subcarrier spacing and symbol length / duration are a function of the parameter set. The subcarrier spacing may be equal to 2 μ × 15kHz, where μ is parameter set 0 to 5. Thus, parameter set μ=0 has a subcarrier spacing of 15kHz, while parameter set μ=5 has a subcarrier spacing of 480kHz. The symbol length / duration is inversely related to the subcarrier spacing. Figure 4A , Figure 4B , Figure 4C and Figure 4D An example is provided for slot configuration 0 with 14 symbols per slot and parameter set μ=2 with 4 slots per subframe. The slot duration is 0.25 ms, the subcarrier spacing is 60 kHz, and the symbol duration is approximately 16.67 μs.

[0086] like Figure 4A , Figure 4B , Figure 4C and Figure 4D As depicted in , a resource grid may be used to represent a frame structure. Each time slot includes a resource block (RB) (also referred to as a physical RB (PRB)) extending, for example, over 12 consecutive subcarriers. The resource grid is divided into a plurality of resource elements (REs). The number of bits carried by each RE depends on the modulation scheme.

[0087] like Figure 4A As illustrated, some of the REs carry data for the UE (eg, Figure 1 and Figure 3 The RS may include a demodulation RS (DMRS) and / or a channel state information reference signal (CSI-RS) for channel estimation at the UE. The RS may also include a beam measurement RS (BRS), a beam refinement RS (BRRS), and / or a phase tracking RS (PT-RS).

[0088] Figure 4B Examples of various DL channels within a subframe of a frame are illustrated.The Physical Downlink Control Channel (PDCCH) carries DCI within one or more Control Channel Elements (CCEs), each CCE comprising, for example, nine RE Groups (REGs), each REG comprising, for example, four consecutive REs in an OFDM symbol.

[0089] The primary synchronization signal (PSS) may be in symbol 2 of a particular subframe of a frame. The PSS is received by a UE (e.g., Figure 1 and Figure 3 104) is used to determine subframe / symbol timing and physical layer identification.

[0090] The Secondary Synchronization Signal (SSS) may be in symbol 4 of a specific subframe of a frame. The SSS is used by the UE to determine the physical layer cell identity group number and the radio frame timing.

[0091] Based on the physical layer identifier and the physical layer cell identifier group number, the UE can determine the physical cell identifier (PCI). Based on the PCI, the UE can determine the location of the aforementioned DMRS. The physical broadcast channel (PBCH) carrying the master information block (MIB) can be logically grouped with the PSS and SSS to form a synchronization signal (SS) / PBCH block. The MIB provides the number of RBs in the system bandwidth and the system frame number (SFN). The physical downlink shared channel (PDSCH) carries user data, broadcast system information (such as system information blocks (SIBs)) that is not sent via the PBCH, and / or paging messages.

[0092] like Figure 4C As illustrated, some REs carry DMRS for channel estimation at the base station (indicated as R for a particular configuration, but other DMRS configurations are also possible). The UE may send DMRS for PUCCH and DMRS for PUSCH. The PUSCH DMRS may be sent, for example, in the first one or two symbols of the PUSCH. The PUCCH DMRS may be sent in different configurations depending on whether a short PUCCH or a long PUCCH is sent and depending on the specific PUCCH format used. UE 104 may send a sounding reference signal (SRS). The SRS may be sent, for example, in the last symbol of a subframe. The SRS may have a comb structure, and the UE may send the SRS on one of the comb structures in the comb structure. The SRS may be used by the base station for channel quality estimation to achieve frequency-dependent scheduling of the UL.

[0093] Figure 4D Examples of various UL channels within a subframe of a frame are illustrated. The PUCCH may be located as indicated in one configuration. The PUCCH carries uplink control information (UCI), such as scheduling requests, channel quality indicator (CQI), precoding matrix indicator (PMI), rank indicator (RI), and HARQ ACK / NACK feedback. The PUSCH carries data and may additionally be used to carry a buffer status report (BSR), a power headroom report (PHR), and / or UCI.

[0094] Example Beam Refinement Procedure

[0095] In mmWave systems, beamforming may be required to overcome high path loss. As described herein, beamforming may refer to establishing a link between a BS and a UE, where the two devices form beams corresponding to each other. Both the BS and the UE find at least one sufficient beam to form a communication link. The BS beam and the UE beam form a so-called beam pair link (BPL). As an example, on the DL, the BS may use a transmit beam, and the UE may receive a transmission using a receive beam corresponding to the transmit beam. The combination of a transmit beam and a corresponding receive beam may be a BPL.

[0096] As part of beam management, the beams used by the BS and UE must be refined from time to time due to changing channel conditions (e.g., due to movement of the UE or other objects). In addition, the performance of the BPL may experience fading due to Doppler spread. Due to the changing channel conditions over time, the BPL should be updated or refined periodically. Therefore, it may be beneficial if the BS and UE monitor the beams and the new BPL.

[0097] At least one BPL must be established for network access. As mentioned above, new BPLs may need to be discovered later for different purposes. The network may decide to use different BPLs for different channels, or for communicating with different BSs (TRPs), or as a fallback BPL in the event of failure of the existing BPL.

[0098] The UE usually monitors the quality of the BPL, and the network may refine the BPL from time to time.

[0099] Figure 5 An example 500 for BPL discovery and refinement is illustrated. In 5G-NR, P1, P2, and P3 procedures are used for BPL discovery and refinement. The network uses the P1 procedure to enable discovery of new BPLs. In the P1 procedure, Figure 5 As shown, the BS transmits different symbols of the reference signal, and each beam is formed in a different spatial direction so as to reach several (most, all) relevant positions of the cell. In other words, the BS transmits beams in different directions over time using different transmit beams.

[0100] In order to successfully receive at least one symbol of this "P1 signal", the UE must find a suitable receive beam. It searches using the available receive beams and applying a different UE beam during each occurrence of the periodic P1 signal.

[0101] Once the UE successfully receives a symbol of the P1 signal, it has discovered the BPL. The UE may not want to wait until it has found the best UE receive beam, as this may delay subsequent actions. The UE may measure the reference signal received power (RSRP) and report the symbol index together with the RSRP to the BS. Such a report will typically contain the discovery of one or more BPLs.

[0102] In one example, the UE may determine a received signal with a high RSRP. The UE may not know which beam the BS used to transmit; however, the UE may report to the BS the time when it observed a signal with a high RSRP. The BS may receive the report and may determine which BS beam the BS used at a given time.

[0103] The BS may then provide P2 and P3 procedures to refine the individual BPLs. The P2 procedure refines the BS beam of the BPL. The BS may transmit several symbols of the reference signal using a different BS beam that is spatially close to the BS beam of the BPL (the BS performs the sweep using neighboring beams around the selected beam). In P2, the UE keeps its beam constant. Therefore, when the UE uses the same beam as in the BPL (e.g. Figure 5 The BS beams used for P2 may be different from the BS beams used for P1 in that they may be spaced closer together or they may be more focused. The UE may measure the RSRP of the various BS beams and indicate the best RSRP to the BS.

[0104] The P3 procedure refines the UE beam for BPL (see Figure 5 3 procedure in BPL). When the BS beam remains constant, the UE uses different receive beams for scanning (UE uses adjacent beams to perform sweeping). The UE can measure the RSRP of each beam and identify the best UE beam. Afterwards, the UE can use the best UE-beam of BPL and report the RSRP to the BS.

[0105] After the timeout, the BS and UE establish several BPLs. When the BS sends a specific channel or signal, it lets the UE know which BPL will be involved so that the UE can tune in the direction of the correct UE receive beam before the signal starts. In this way, each sample of the signal or channel can be received by the UE using the correct receive beam. In one example, the BS can indicate which BPL is involved for the scheduled signal (SRS, CSI-RS) or channel (PDSCH, PDCCH, PUSCH, PUCCH). In NR, this information is called QCL indication.

[0106] Two antenna ports are QCL if the properties of the channel over which symbols on one antenna port are communicated can be inferred from the channel over which symbols on the other antenna port are communicated. QCL supports at least beam management functionality, frequency / timing offset estimation functionality, and RRM management functionality.

[0107] The BS can use the BPL that the UE has received in the past. The transmit beams for the signal to be sent and the previously received signal are both pointed in the same direction or are QCL. The UE may require a QCL indication (before the signal to be received) so that the UE can use the correct receive beam for each signal or channel. When the BPL used for the signal or channel changes, some QCL indications may sometimes be required, and some QCL indications are required for each scheduled instance. The QCL indication can be sent in the downlink control information (DCI), which can be part of the PDCCH channel. Because DCI is needed to control information, it may be expected that the number of bits required to indicate the QCL is not too large. The QCL can be sent in a medium access control-control element (MAC-CE) or a radio resource control (RRC) message.

[0108] According to an example, whenever a UE reports a BS beam that it has received with sufficient RSRP and the BS decides to use that BPL in the future, the BS assigns it a BPL tag. Thus, two BPLs with different BS beams can be associated with different BPL tags. BPLs based on the same BS beam can be associated with the same BPL tag. Thus, according to this example, the tag is a function of the BS beam of the BPL.

[0109] As mentioned above, wireless systems such as millimeter wave (mmW) systems are bringing gigabit speeds to cellular networks due to the availability of large amounts of bandwidth. However, the unique challenges of severe path loss faced by such wireless systems require new technologies such as hybrid beamforming (analog and digital) that were not present in 3G and 4G systems. Hybrid beamforming can enhance the link budget / signal-to-noise ratio (SNR) available during RACH.

[0110] In such a system, a Node B (NB) and a user equipment (UE) can communicate via an active beamforming transmit beam. An active beam can be considered as a paired transmit (Tx) and receive (Rx) beam between the NB and the UE, which carries data and control channels such as PDSCH, PDCCH, PUSCH, and PUCCH. As described above, the transmit beam used by the NB for downlink transmission and the corresponding receive beam used by the UE for downlink transmission can be referred to as a beam pair link (BPL). Similarly, the transmit beam used by the UE for uplink transmission and the corresponding receive beam used by the NB for uplink transmission can also be referred to as a BPL.

[0111] In such a system, a Node B (NB) and a user equipment (UE) can communicate via an active beamforming transmit beam. An active beam can be considered as a paired transmit (Tx) and receive (Rx) beam between the NB and the UE, which carries data and control channels such as PDSCH, PDCCH, PUSCH, and PUCCH. As described above, the transmit beam used by the NB for downlink transmission and the corresponding receive beam used by the UE for downlink transmission can be referred to as a beam pair link (BPL). Similarly, the transmit beam used by the UE for uplink transmission and the corresponding receive beam used by the NB for uplink transmission can also be referred to as a BPL.

[0112] Since the direction of the reference signal is unknown to the UE, the UE may need to evaluate several beams to obtain the best Rx beam for a given NB Tx beam. However, if the UE must "sweep" through all of its Rx beams to perform measurements (e.g., to determine the best Rx beam for a given NB Tx beam), the UE may incur significant measurement delays and battery life impacts. In addition, having to sweep through all Rx beams is extremely resource inefficient. Therefore, various aspects of the present disclosure provide techniques to assist UEs in performing measurements of serving and neighboring cells when using Rx beamforming.

[0113] Example Beam Management

[0114] In wireless communications, various procedures for beam management may be performed. Figure 6 602, the network may sweep several beams, for example, via synchronization signal blocks (SSBs), as described herein with respect to Figure 4B Further described. The network may configure the UE to have a random access channel (RACH) resource associated with a beamformed SSB to facilitate initial access via the RACH resource. In certain aspects, the SSB may have a wider beam shape compared to other reference signals such as a channel state information reference signal (CSI-RS). The UE may use SSB detection to identify a RACH opportunity (RO) for transmitting a RACH preamble (e.g., as part of a contention CBRA procedure).

[0115] In connected mode 604, the network and the UE may perform hierarchical beam refinement, which includes beam selection (e.g., a process referred to as P1), beam refinement for the transmitter (e.g., a process referred to as P2), and beam refinement for the receiver (e.g., a process referred to as P3). For example, in beam selection (P1), the network may sweep the beams and the UE may report the beam with the best channel properties. In beam refinement for the transmitter (P2), the network may sweep narrower beams and the UE may report the beam with the best channel properties among the narrow beams. In beam refinement for the receiver (P3), the network may repeatedly use the same beam for transmission, and the UE may refine the spatial reception parameters (e.g., spatial filters) used to receive signals from the network via the beam. In certain aspects, the network and the UE may perform complementary procedures (e.g., U1, U2, and U3) for uplink beam management.

[0116] In certain cases where a beam failure occurs (e.g., due to beam misalignment and / or obstruction), the UE may perform a beam failure recovery (BFR) procedure 606, which may allow the UE to return to connected mode 604 without performing a radio link failure procedure 608. For example, the UE may be configured with candidate beams for beam failure recovery. In response to detecting a beam failure, the UE may request the network to perform beam failure recovery via one of the candidate beams (e.g., one of the candidate beams having a reference signal received power (RSRP) above a certain threshold). In certain cases where a radio link failure (RLF) occurs, the UE may perform an RLF procedure 608 to recover from the radio link failure, such as a RACH procedure.

[0117] AI / ML functional framework for RAN intelligence

[0118] The AI / ML functional framework includes a data collection function 702, a model training function 704, a model inference function 706, and an actor function 708, which interoperate to provide a platform for collaboratively applying AI / ML to various procedures in the RAN.

[0119] The data collection function 702 typically provides input data to the model training function 704 and the model inference function 706. AI / ML algorithm specific data preparation (e.g., data preprocessing and cleaning, formatting and transformation) may not be performed in the data collection function 702.

[0120] Examples of input data to the data collection function 702 (or other functions) may include measurements from UEs or different network entities, feedback from auto-author functions, and outputs from AI / ML models. In some cases, data analysis required at the model training function 704 and the model inference function 706 may be performed at the data collection function 702. As shown, the data collection function 702 may deliver training data to the model training function 704 and deliver inference data to the model inference function 706.

[0121] The model training function 704 can perform AI / ML model training, validation, and testing, which can generate model performance metrics as part of the model testing procedure. If necessary, the model training function 704 can also be responsible for data preparation (e.g., data preprocessing and cleaning, formatting, and transformation) based on the training data delivered by the data collection function 702.

[0122] The model training function 704 may provide model deployment / update data to the model interface function 706. The model deployment / update data may be used to initially deploy a trained, validated, and tested AI / ML model to the model inference function 706 or to deliver an updated model to the model inference function 706.

[0123] As shown, the model inference function 706 can provide AI / ML model inference output (e.g., prediction or decision) to the actor function 708 and sometimes also provide model performance feedback to the model training function 704. The model inference function 706 can also sometimes be responsible for data preparation (e.g., data preprocessing and cleaning, formatting and transformation) based on the inference data delivered by the data collection function 702.

[0124] Inference output of the AI / ML model may be produced by the model inference function 706. The specific details of this output may be specific in terms of the use case. Sometimes, model performance feedback may be used to monitor the performance of the AI / ML model. In some cases, for example, if certain information derived from the model inference function is applicable to improving the AI / ML model trained in the model training function 704, the model performance feedback may be delivered to the model training function 704.

[0125] The model inference function 706 may signal the outputs of the model to nodes that have requested them (e.g., via subscriptions) or to nodes that take actions based on the outputs from the model inference function. Prior to deployment, the AI / ML models used in the model inference function 706 may require initial training, validation, and testing by the model training function. The model training function 704 and the model inference function 706 may be able to request specific information for training or executing the AI / ML algorithm and avoid receiving unnecessary information. The nature of such information may depend on the use case and the AI / ML algorithm.

[0126] The actor function 708 may receive output from the model inference function 706, which may trigger or perform a corresponding action. The actor function 708 may trigger actions for other entities or for itself. The feedback generated by the actor function 708 may provide information for deriving training data, inferring data, or for monitoring the performance of the AI / ML model. As described above, the input data for the data collection function 702 may include this feedback from the author function 708. Feedback from the author function 708 or other network entities (via the data collection function) may also be used at the model inference function 706.

[0127] The AI / ML functional framework 700 can be deployed in various RAN intelligence-based use cases. Such use cases may include CSI feedback enhancement, enhanced beam management (BM), positioning and location (Pos-Loc) accuracy enhancement, and various other use cases.

[0128] In certain aspects, the UE or BS may use continuously measured or reported L1-RSRP in the time domain to perform ML-based beam prediction. In some cases, a pre-trained deep neural network (DNN) model may be used for such ML-based predictive beam management.

[0129] Traditionally, beam quality and failure are identified through measurement reports carried by relevant downlink (DL) and uplink (UL) reference signals (e.g., SSB, CSI-RS, RSRP), which increases beam selection latency and beam management overhead, while in the same case, beam selection accuracy may be limited due to power and overhead constraints, which may lead to poor system performance.

[0130] In contrast, AI / ML-based predictive beam management can reduce the amount of reference signal transmission used to predict unmeasured beam quality and the future possibility of beam blocking / failure. In predictive beam management, beam prediction can be a highly nonlinear problem that can be effectively solved by a pre-trained DNN model that can predict future beam quality, for example, based on UE mobility speed and trajectory that are difficult to model through conventional statistical processing methods.

[0131] Aspects related to ML model management and auxiliary information

[0132] Conventionally, several techniques have been used to help determine the channel state between wireless communication devices so that those devices can optimize their wireless communication configurations (e.g., select the best beam for sending and receiving data). For example, a channel state information reference signal (CSI-RS)) can be sent by one device and measured by another device in order to estimate the channel state and provide channel state information (CSI) feedback useful for optimizing wireless communication between the two devices.

[0133] However, due to the increasing complexity and capabilities of wireless communication devices, such as those capable of transmitting and receiving through multiple input and output antenna ports (e.g., implementing multiple-input multiple-output (MIMO) technology), conventional techniques may require significant processing power and time, which reduces the performance of both the device and the entire wireless communication network. These technical problems are exacerbated by typical use cases and environments for wireless communication, which are often dynamic. In other words, because channel states change frequently, channel estimation and feedback procedures are often performed frequently, resulting in high power usage and significant network overhead for wireless communication systems (e.g., in terms of time and frequency resources dedicated to channel estimation). One way to alleviate such problems is to implement a machine learning model that can perform various functions related to channel state estimation and feedback more accurately and efficiently.

[0134] For example, conventional wireless communication systems may use, for example, time division multiplexing (TDM), code division multiplexing (CDM), and / or frequency division multiplexing (FDM) to allocate N packets per resource block. t Multiplex N resource elements t Such a system may typically achieve a resource block density between 0.5 and 1, such that a resource element is transmitted in every other or every single resource block. In contrast, a machine learning model deployed by a transmitting device (e.g., a base station) may be trained to perform N on L resource elements per resource block. t Multiplexing of ports, where L <N t , which therefore reduces the number of resource elements required for channel estimation, leaving more resource elements available for data transmission. In addition to reducing the number of resource elements required for channel estimation (which reduces power and enhances resource utilization), such models can also operate at reduced resource block density (e.g., below 0.5) and can also implement non-uniform resource block patterns, which further improves the aforementioned benefits. On the receiving device (e.g., user equipment) side, a machine learning-based channel estimator can be trained to recover the complete channel on all resource blocks (e.g., N t ports), while receiving a reduced number of resources L. In various aspects, the CSI-RS multiplexing models at the transmitter side and the receiver side may be trained jointly or sequentially.

[0135] As another example, conventional CSI reporting configurations may rely on a precoding matrix indicator (PMI) search algorithm and a PMI codebook to determine the best PMI codeword (e.g., CSI feedback) and report it to the network. However, machine learning-based models (such as encoders and decoders) may be trained to directly generate CSI feedback, which eliminates the need for a PMI search algorithm (replaced by the encoder) and a PMI codebook (replaced by the decoder). In various aspects described herein, a CSI encoder at the user equipment side may be trained to compress a channel estimate to a few bits that are then reported to a network entity (e.g., a base station), while a CSI decoder at the network entity side is trained to use the reported bits to recover the channel or precoding matrix.

[0136] Thus, in general, machine learning models can be trained to perform many functions related to channel estimation and feedback, and such models can often be more accurate, faster, more power efficient, and better able to maintain performance in very dynamic radio environments. However, it is still important to monitor the performance of such machine learning models to ensure robust performance over time.

[0137] Figure 8 Depicted is an example 800 of monitoring machine learning model performance over time.

[0138] As shown, during time interval 802, model output 804 closely tracks actual value (e.g., of channel estimate) 806. The model may be deployed by user equipment, such as with respect to Figure 1 and Figure 3 804. During interval 802, there are two reporting events where the device deploying the model reports the performance of the model, including two out-of-distribution (OOD) events. Notably, OOD events depict situations where the model output deviates significantly (e.g., based on a threshold) from the actual value 806. In these situations, the trained model may be processing input data that is significantly different from the training data used to train the model, and therefore the model output has become unreliable. Based on these OOD events, such as regarding Figures 1 to 3 A network entity such as base station 102 is described as being able to determine the delivery model update.

[0139] The second time interval 804 demonstrates various possible outcomes. Without the model update, the original model output 804 deviates significantly from the actual value 806. In contrast, the updated model output 808 again closely tracks the actual value 806. Furthermore, fallback methods such as conventional non-machine learning based methods are depicted to demonstrate that such methods may be better than a poorly performing machine learning model, but worse than a well performing machine learning model.

[0140] There are various options for what type of information is fed back as channel state feedback (CSF), and the decision on what to feed back may depend on the specific implementation (e.g., whether based on AI / ML). The type of feedback can vary on a continuum from relatively sparse information such as rank indicator (RI), PMI, and channel quality indicator (CQI) to detailed full channel information.

[0141] In current systems, for CSI feedback, the CSI reporting configuration usually includes a codebook. The codebook is used as a Precoding Matrix Indicator (PMI) dictionary, from which the UE can select and report the best PMI codeword, thereby reporting the PMI using a bit sequence.

[0142] like Fig. 9 As shown in the example block diagram 900 of , in some cases, the AI-based CSI feedback may be replaced by a CSI encoder 902 (at the UE) and a decoder 9004 at the network (e.g., gNB) to replace the codebook. In some cases, the encoder may not be required and only the decoder may be used. Typically, the encoder is similar to the PMI search algorithm in current systems, and the decoder is typically similar to the PMI codebook, which is used to convert the CSI report bits into PMI codewords.

[0143] like Fig.10 As shown in Table 1000, the input to the decoder includes a downlink channel matrix (H), a downlink pre-decoder (V), and an interference covariance matrix (Rnn). The output of the decoder can be a downlink channel matrix (H), a transmit covariance matrix, a downlink pre-decoder (V), and an interference covariance matrix (Rnn), where H or V can correspond to the original channel or a channel pre-whitened by the UE based on its demodulation filter.

[0144] As about Figure 8 As described, in the case of OOD or model failure, fallback (to another algorithm / model) or retraining can be performed, or some other action can be taken to adapt to the new scenario (or cell condition) that caused the failure. One potential design decision is which entity (UE or network entity) should perform model monitoring and which entity should make decisions about model retraining or switching.

[0145] Various aspects of the present disclosure provide various options for ML model performance monitoring and decision making. Various aspects provide various options, including a first option involving network-side (e.g., gNB-side) ML model performance monitoring and a second option involving UE-side ML model performance monitoring. The monitored ML model can be running on the UE or the gNB.

[0146] For the first option, in case the monitoring is performed by an entity on the network side, the decision can be made on the UE or network side, with or without auxiliary information from the UE side. According to the second option, in case the monitoring is performed on the UE side, the UE can make the decision with or without auxiliary information from the network (gNB side). As used herein, the term "UE side" may refer to the UE or an entity associated with the UE, such as a UE server (communicating with the UE), a UE vendor, or a model designer or a model repository. Similarly, the term "network side" or "gNB side" may refer to a network entity (such as a gNB, a gNB server) or any other element at the network side (such as a model repository).

[0147] Fig.11 A call flow diagram for network-side performance monitoring is illustrated, where retraining / switching decisions are made at the network side.

[0148] As shown, the gNB may request assistance information from the UE (step 0), and in response, the UE may report assistance information (step 1). Based on the monitoring, (at step 2) the gNB may generate one or more key performance indicators (KPIs) and generate a monitoring performance report (e.g., including the KPIs). The gNB sends the monitoring performance report to the UE (at step 3). As shown, in some cases (at step 4), the UE may forward the report to an entity associated with the UE (e.g., a UE vendor).

[0149] If a condition is detected based on monitoring, the gNB may decide to retrain or switch the model. In this case, the gNB may transmit (at step 5) a model deactivation command along with an indication that the model is to be retrained or switched. In response, the UE may transmit a model deactivation retrain / switch request to the UE vendor (at step 6).

[0150] In some cases, the UE may also transmit an acknowledgment to the gNB (at step 7) that the UE received the model deactivation and retraining / handover request. Finally, the entity may engage in a data collection and retraining procedure (at step 8). Exactly how this procedure is performed may depend on multiple factors, such as the type of model, whether the model is being switched or retrained, and where the model is running.

[0151] Fig.12 A call flow diagram for network-side performance monitoring is illustrated, where the retraining / handover decision is made at the UE side (rather than the network side).

[0152] As shown in the figure, steps 0 to 4 can be Fig.11, where the retraining / handover decision is made at the network side. However, steps 5 to 7 may be different, in which case the UE vendor transmits a model deactivation retraining / handover request to the UE (step 5). In response, the UE may forward the model deactivation retraining / handover request to the gNB (step 6). The gNB may then confirm the request (at step 7). Finally, the entity may engage in data collection and retraining procedures (at step 8). In some other instances, the UE may make a model retraining / handover decision directly without the need for a UE server. In this case, the UE may transmit a model deactivation retraining / handover request to the gNB. In addition, the UE may initiate a model deactivation retraining / handover and data collection procedure with its UE server.

[0153] like Fig.11 and Fig.12 As shown, in some cases, auxiliary information from the UE (steps 1 and 0) is used to generate a model performance report. The type and content of the auxiliary information may vary. For example, in some cases, the UE may provide the gNB with ground truth (e.g., V_ideal based on actual data collected "on the ground", as opposed to predicted), the corresponding model ID used in the inference, and the corresponding CSI report configuration and its triggering timing or reporting timing. As another example, the gNB may transmit a CSI report trigger in step 0 to request regular CSI feedback using a non-AI codebook (such as Type I, Type II, or eType II). In step 1, the UE may report the corresponding PMI based on the CSI report triggered in step 0.

[0154] In some cases, the gNB may compare the CSI inference (prediction) results (i.e., CSI decoder output) with the ground truth (provided in the auxiliary information). The comparison may be done in terms of KPI (e.g., mean square error MSE), spectral efficiency, throughput, or cosine similarity. In some other cases, the auxiliary information is a conventional non-AI-based (or non-ML-based) CSI report, where the PMI is reported using a non-AI / ML codebook such as Type I, Type II, or eType II. The conventional non-AI-based PMI report is based on the same CSI-RS resources used for channel measurement as the CSI inference results. The gNB may compare the CSI inference (prediction) results with the non-AI-based PMI. The comparison may be done in terms of KPI (e.g., mean square error MSE), spectral efficiency, throughput.

[0155] In some cases, the assistance information may be reported on request (from the gNB at step 0). In other cases, the assistance information may be reported periodically (P) or semi-persistently (SP). As described above, in some cases, the gNB may generate performance reports without assistance information. In some cases, the gNB may use SRS measurements as ground truth to determine OOD / model failure. In some cases, the gNB may have limited information (e.g., {z, Vhat}) and may use AI / ML-based methods to determine whether {z, Vhat} is in distribution or OOD with respect to {z_training, Vhat_training} (e.g., using a NN model separate from the model being monitored or based on statistics inside the decoder). Here, z and z_training represent potential information output by the CSI encoder and input to the CSI decoder, while Vhat and Vhat_training represent the CSI decoder output.

[0156] In some cases, the reporting information in the performance report may indicate a performance gap with the ground truth or may indicate general system performance. In some cases, the reporting information may include statistics for each individual inference or each multiple inferences. The reporting information may also include the likelihood of adaptation in each existing model (e.g., {0.5, 0.3, 0.2}), indicating the likelihood of adaptation in one of the three models 1, 2, and 3, respectively. Generating this information may require a separate NN running at the gNB side, which outputs the likelihood. The performance report may be transmitted based on an event trigger, may be transmitted periodically, or may be transmitted semi-persistently.

[0157] The decision to retrain / switch the model at the gNB or UE side (e.g., at steps 5 to 6) may be based on the monitoring and the generated performance reports. Based on these, the gNB or UE server may decide whether to stay on the current model (in which case, additional steps 5 to 7 may not be taken), or may determine that the current model has failed and transmit a model deactivation request.

[0158] In the case where the failed sample is adapted in other existing models, a model switching decision can be made (e.g., in addition to model deactivation, the model is updated with the corresponding ID based on the adaptation). Otherwise (e.g., if it is not adapted), the entity making the decision can transmit a model retraining request (e.g., in addition to model deactivation).

[0159] For the decision at the gNB side, if the handover is determined, the signaling may also include information of the model (e.g., KPI), for example, with an updated model ID. If the decision is to retrain, the signaling may also include information about the failed samples (e.g., delay spread, average delay, average gain, Doppler, spatial information such as angle of arrival - AoA, angle of departure - AoD, zenith angle of arrival - ZoA, and zenith angle of departure - ZoD). The signaling mechanism for steps 5 and 6 may be the same as the signaling mechanism for steps 3 and 4 (e.g., RRC or MAC layer signaling). The signaling for steps 5 and 6 may be via the same signaling for steps 3 and 4.

[0160] For the decision at the UE side, after transmitting the initial deactivation request, the UE side may refrain from transmitting subsequent deactivation requests for a timer period. For deactivation confirmation, at step 7, for the gNB side decision, the UE may transmit an ACK for the deactivation command, indicating successful receipt of the command and starting retraining / handover. For the UE side decision, the gNB may transmit an ACK via dedicated signaling or via RRC reconfiguration of the AI / ML model (in other words, the UE may consider the RRC reconfiguration as an ACK for receiving the handover request). Fig.13 A call flow diagram for UE-side performance monitoring is illustrated. As described above, in this case, a retraining / handover decision can also be made at the UE side.

[0161] As shown, step 0 may include two steps, where the UE provider transmits a request for assistance information (at step 0.1), and the UE may forward the request to the gNB (at step 0.2). For step 1, the gNB may transmit a provision for assistance information at step 1.1, and the UE may forward the provision to the UE provider (at step 1.2). At step 2, the UE provider may generate one or more KPIs and generate a monitoring performance report (e.g., including the KPIs). The UE provider sends the monitoring performance report to the UE (at step 3). In some cases, the UE may send the assistance information request directly to the gNB without involving the UE provider. In some cases, once the UE receives the assistance information from the gNB, the UE may use the assistance information to generate one or more KPIs and generate a monitoring performance report (e.g., including the KPIs).

[0162] As shown, based on the monitoring performance report, the UE may decide to retrain or switch models. In this case, (at step 4) the UE may transmit a model deactivation command to the gNB along with an indication that the model will be retrained or switched. The gNB may confirm the request (at step 5). Finally, the entity may engage in a data collection and retraining procedure (at step 6).

[0163] For UE-side performance monitoring, the assistance information from the gNB (at steps 0 and 1) can be used to generate performance monitoring reports. In some cases, the gNB can provide the UE with the inference result (e.g., CSI decoder output), the corresponding model ID used in the inference, the corresponding CSI report configuration, and its triggering timing or reporting timing. In this case, the UE can compare the CSI inference result (i.e., CSI decoder output) with the ground truth or compare it with the baseline CSI codebook. The comparison can be done in terms of KPIs (e.g., MSE, spectral efficiency, throughput, or cosine similarity).

[0164] The assistance information report may be transmitted from the UE side on request (step 0), periodically, or semi-persistently. In some cases, the UE side may be able to generate a performance report without assistance information. If the UE has a CSI decoder, the UE is able to compare the inference result with the ground truth or baseline CSI codebook. If the UE has limited information (such as only {H, Z}), the UE may use an AI / ML-based approach to determine whether {H, z} is in distribution or OOD with respect to {H_training, z_training} (e.g., using a separate neural network-based model or based on underlying statistics inside the CSI encoder).

[0165] The reporting information (at steps 2 and 3) may include the performance gap with the ground truth, the general system performance, or the statistics per individual inference or per multiple inferences. In some cases, the reporting information may include the likelihood of fitting in each existing model (e.g., {0.5, 0.3, 0.2} likelihoods for three candidate models 1, 2, 3, respectively). This method may involve a separate model (e.g., NN) running at the gNB side, which may be configured to output the likelihood information. In some cases, the performance report may be transmitted based on an event trigger, may be transmitted periodically, or may be transmitted semi-persistently.

[0166] UE-side decisions about model retraining / switching can be based on its own monitoring and generated performance reports. Based on this information, the UE vendor (e.g., UE server) can decide whether to stay on the current model (where steps 5 to 7 are not performed) or whether the current model is invalid (e.g., resulting in the transmission of a model deactivation request). In some cases, if the invalidated samples are adapted in another existing model, a model switching decision can be made, and the model switching request can include an indication of the updated model, such as with a corresponding model ID in addition to the model deactivation. If no adaptation is determined, the UE (UE vendor) can transmit a model retraining request (e.g., in addition to the model deactivation). After transmitting the first deactivation request, the UE can refrain from transmitting a second deactivation request, for example, for a timer period.

[0167] In some other cases, the UE or UE vendor can transmit performance reports to the gNB side. The gNB can make decisions on model deactivation, retraining / switching. The gNB can then transmit a command to the UE to deactivate the current model. Model retraining / switching and data collection can be initiated by the UE side or the gNB side.

[0168] The deactivation confirmation (step 7) may be transmitted in any suitable type of signaling. For example, the confirmation may be transmitted via dedicated signaling or (implicitly) via RRC reconfiguration of the AI / ML model.

[0169] Example Operation of User Equipment

[0170] Fig.14 The method for use by a UE such as Figure 1 and Figure 3 An example of a method 1400 for performing wireless communications with a UE 104).

[0171] Method 1400 begins at step 1405, where a performance report for an ML model running on at least one of a UE or a network entity is obtained from a network entity. In some cases, the operation of this step refers to the same as in reference to Fig.18 The described circuits for obtaining and / or codes for obtaining may be or may be performed by the circuits for obtaining and / or codes for obtaining.

[0172] Then, method 1400 proceeds to step 1410, where changes to the ML model are made based on the performance report. In some cases, the operations of this step refer to the Fig.18 The described circuits for participating and / or code for participating may be or may be executed by the circuits for participating and / or code for participating.

[0173] In some aspects, the method 1400 further includes: forwarding the performance report to an entity associated with the UE. In some cases, the operation of this step refers to the operation of Fig.18 The described circuits for forwarding and / or codes for forwarding may be or may be performed by the circuits for forwarding and / or codes for forwarding.

[0174] In some aspects, the method 1400 further includes: sending assistance information to the network entity to assist the network entity in generating the performance report. In some cases, the operation of this step refers to the operation of Fig.18 The described circuits for transmitting and / or codes for transmitting may be or may be performed by circuits for transmitting and / or codes for transmitting.

[0175] In some aspects, the assistance information includes at least one of: ground truth information collected by the UE, an ID of the ML model, a corresponding CSI reporting configuration, and information about a CSI reporting configuration triggering occasion or a reporting occasion.

[0176] In some aspects, the assistance information is sent in response to a request from a network entity, or periodically, or semi-persistently.

[0177] In some aspects, the performance report indicates at least one of: an indication of a difference between predicted and actual performance, general system performance, a statistic for each individual inference made by the ML model, or a statistic for each plurality of inferences made by the ML model.

[0178] In some aspects, the performance report indicates the likelihood of fitting in different ML models.

[0179] In some aspects, engaging in changes to the ML model based on the performance report includes engaging in retraining of the ML model or switching to a different ML model.

[0180] In some aspects, the method 1400 further includes: receiving an instruction from the network entity to retrain the ML model or switch to a different ML model. In some cases, the operation of this step refers to the operation of Fig.18 The described circuits for receiving and / or codes for receiving may be or may be performed by the circuits for receiving and / or codes for receiving.

[0181] In some aspects, the indication includes deactivating the ML model.

[0182] In some aspects, the method 1400 further includes: forwarding the indication to an entity associated with the UE. In some cases, the operation of this step refers to the operation of Fig.18 The described circuits for forwarding and / or codes for forwarding may be or may be performed by the circuits for forwarding and / or codes for forwarding.

[0183] In some aspects, where the indication is to retrain the ML model, the indication also includes information for retraining.

[0184] In some aspects, where the indication is to switch to a different ML model, the indication also includes an identification of the different ML model.

[0185] In some aspects, the method 1400 further includes: sending a confirmation of the receipt of the indication to the network entity. In some cases, the operation of this step refers to as referring to Fig.18 The described circuits for transmitting and / or code for transmitting may be or may be performed by the circuits for transmitting and / or code for transmitting.

[0186] In some aspects, the method 1400 further includes: sending an instruction to the network entity to retrain the ML model or switch to a different ML model. In some cases, the operation of this step refers to the operation of Fig.18 The described circuits for transmitting and / or codes for transmitting may be or may be performed by circuits for transmitting and / or codes for transmitting.

[0187] In some aspects, the indication includes deactivating the ML model.

[0188] In some aspects, the method 1400 further includes: before sending the indication to the network entity, receiving an indication from an entity associated with the UE to retrain the current ML model or switch to a different ML model. In some cases, the operation of this step refers to as described in reference Fig.18 The described circuits for receiving and / or codes for receiving may be or may be performed by the circuits for receiving and / or codes for receiving.

[0189] In some aspects, where the indication is to retrain the ML model, the indication also includes information for retraining.

[0190] In some aspects, where the indication is to switch to a different ML model, the indication also includes an identification of the different ML model.

[0191] In one aspect, method 1400 or any aspect related thereto may be performed by an apparatus such as Fig.18 The method 1400 is performed by a communication device 1800 that includes various components operable, configured, or adapted to perform the method 1400. The communication device 1800 is described in more detail below.

[0192] It should be noted that Fig.14 This is merely one example of a method, and other methods including fewer, additional, or alternative steps may also be consistent with the present disclosure.

[0193] Fig.15 The method for use by a UE such as Figure 1 and Figure 3 An example of a method 1500 for performing wireless communications with a UE 104).

[0194] Method 1500 begins at step 1505, where a performance report for an ML model running on at least one of a UE or a network entity is generated. In some cases, the operation of this step refers to the same as in reference to Fig.18 The described circuits for generating and / or codes for generating may be or may be executed by the circuits for generating and / or codes for generating.

[0195] Then, method 1500 proceeds to step 1510, where changes to the ML model are made based on the performance report. In some cases, the operations of this step refer to the Fig.18 The described circuits for participating and / or code for participating may be or may be executed by the circuits for participating and / or code for participating.

[0196] In some aspects, method 1500 further includes: forwarding the performance report to a network entity. In some cases, the operation of this step is referred to as reference Fig.18 The described circuits for forwarding and / or codes for forwarding may be or may be performed by the circuits for forwarding and / or codes for forwarding.

[0197] In some aspects, the method 1500 further includes: receiving auxiliary information from a network entity to assist in generating the performance report. In some cases, the operation of this step refers to as referring to Fig.18 The described circuits for receiving and / or codes for receiving may be or may be performed by the circuits for receiving and / or codes for receiving.

[0198] In some aspects, the assistance information includes at least one of: ground truth information collected by the UE, an ID of the ML model, a corresponding CSI reporting configuration, and information about a CSI reporting configuration triggering occasion or a reporting occasion.

[0199] In some aspects, the assistance information is received by the UE or UE provider periodically, or semi-persistently, or in response to a request sent to a network entity.

[0200] In some aspects, the performance report indicates at least one of: an indication of a difference between predicted and actual performance, general system performance, a statistic for each individual inference made by the ML model, or a statistic for each plurality of inferences made by the ML model.

[0201] In some aspects, the performance report indicates the likelihood of fitting in different ML models.

[0202] In some aspects, engaging in changes to the ML model based on the performance report includes engaging in retraining of the ML model or switching to a different ML model.

[0203] In some aspects, method 1500 further includes: sending an instruction to the network entity to retrain the ML model or switch to a different ML model. In some cases, the operation of this step refers to the operation of Fig.18 The described circuits for transmitting and / or code for transmitting may be or may be performed by the circuits for transmitting and / or code for transmitting.

[0204] In some aspects, the method 1500 further includes: before sending the indication to the network entity, receiving an indication from an entity associated with the UE to retrain the ML model or switch to a different ML model. In some cases, the operation of this step refers to as described in reference Fig.18 The described circuits for receiving and / or codes for receiving may be or may be performed by the circuits for receiving and / or codes for receiving.

[0205] In some aspects, where the indication is to retrain the ML model, the indication also includes information for retraining.

[0206] In some aspects, where the indication is to switch to a different ML model, the indication also includes an identification of the different ML model.

[0207] In some aspects, the indication includes deactivating the ML model.

[0208] In some aspects, method 1500 further includes: receiving an instruction from the network entity to retrain the ML model or switch to a different ML model. In some cases, the operation of this step refers to the operation of Fig.18 The described circuits for receiving and / or codes for receiving may be or may be performed by the circuits for receiving and / or codes for receiving.

[0209] In some aspects, the indication includes deactivating the ML model.

[0210] In one aspect, method 1500 or any aspect related thereto may be performed by an apparatus such as Fig.18 The method 1500 is performed by a communication device 1800 that includes various components operable, configured, or adapted to perform the method 1500. The communication device 1800 is described in more detail below.

[0211] It should be noted that Fig.15 This is merely one example of a method, and other methods including fewer, additional, or alternative steps may also be consistent with the present disclosure.

[0212] Example Operations of Network Entities

[0213] Fig.16 shows a method for use by a network entity such as Figure 1 and Figure 3 BS102 or as relative to Figure 2 An example of a method 1600 for wireless communications performed by a decomposed base station) as discussed.

[0214] Method 1600 begins at step 1605, where a performance report for an ML model running on at least one of a UE or a network entity is sent. In some cases, the operation of this step refers to the same as in reference to Fig.18The described circuits for transmitting and / or code for transmitting may be or may be performed by the circuits for transmitting and / or code for transmitting.

[0215] Then, method 1600 proceeds to step 1610, where changes to the ML model are made based on the performance report. In some cases, the operations of this step refer to the Fig.18 The described circuits for participating and / or code for participating may be or may be executed by the circuits for participating and / or code for participating.

[0216] In some aspects, the method 1600 further includes: receiving assistance information generated by the UE; and using the assistance information when generating the performance report. In some cases, the operation of this step refers to as referring to Fig.18 The described circuits for receiving and / or codes for receiving may be or may be performed by the circuits for receiving and / or codes for receiving.

[0217] In some aspects, the assistance information includes at least one of: ground truth information collected by the UE, an ID of the ML model, a corresponding CSI reporting configuration, and information about a CSI reporting configuration triggering occasion or a reporting occasion.

[0218] In some aspects, the assistance information is received in response to a request from a network entity, or periodically, or semi-persistently.

[0219] In some aspects, the performance report indicates at least one of: an indication of a difference between predicted and actual performance, general system performance, a statistic for each individual inference made by the ML model, or a statistic for each plurality of inferences made by the ML model.

[0220] In some aspects, the performance report indicates the likelihood of fitting in different ML models.

[0221] In some aspects, engaging in changes to the ML model based on the performance report includes sending an indication for the UE to retrain the ML model or switch to a different ML model.

[0222] In some aspects, the indication includes deactivating the ML model.

[0223] In some aspects, where the indication is to retrain the ML model, the indication also includes information for retraining.

[0224] In some aspects, where the indication is to switch to a different ML model, the indication also includes an identification of the different ML model.

[0225] In some aspects, the method 1600 further includes: receiving a confirmation that the UE received the indication. In some cases, the operation of this step refers to as referring to Fig.18 The described circuits for receiving and / or codes for receiving may be or may be performed by the circuits for receiving and / or codes for receiving.

[0226] In some aspects, the method 1600 further includes: receiving an indication from the UE to retrain the ML model or switch to a different ML model. In some cases, the operation of this step refers to the operation of Fig.18 The described circuits for receiving and / or codes for receiving may be or may be performed by the circuits for receiving and / or codes for receiving.

[0227] In some aspects, the indication includes deactivating the ML model.

[0228] In one aspect, method 1600 or any aspect related thereto may be performed by an apparatus such as Fig.18 The method 1600 is performed by a communication device 1800 that includes various components operable, configured, or adapted to perform the method 1600. The communication device 1800 is described in more detail below.

[0229] It should be noted that Fig.16 This is merely one example of a method, and other methods including fewer, additional, or alternative steps may also be consistent with the present disclosure.

[0230] Fig.17 shows a method for use by a network entity such as Figure 1 and Figure 3 BS102 or as relative to Figure 2 An example of a method 1700 for wireless communications performed by a decomposed base station) as discussed.

[0231] Method 1700 begins at step 1705, where a performance report for an ML model running on at least one of a UE or a network entity is received. In some cases, the operation of this step refers to the same as in reference to Fig.18 The described circuits for receiving and / or codes for receiving may be or may be performed by the circuits for receiving and / or codes for receiving.

[0232] Then, method 1700 proceeds to step 1710, where changes to the ML model are made based on the performance report. In some cases, the operations of this step refer to the Fig.18 The described circuits for participating and / or code for participating may be or may be executed by the circuits for participating and / or code for participating.

[0233] In some aspects, the method 1700 further includes: sending assistance information to assist the UE in generating the performance report. In some cases, the operation of this step refers to the operation of Fig.18The described circuits for transmitting and / or codes for transmitting may be or may be performed by circuits for transmitting and / or codes for transmitting.

[0234] In some aspects, the assistance information includes at least one of: ground truth information collected by the UE, an ID of the ML model, a corresponding CSI reporting configuration, and information about a CSI reporting configuration triggering occasion or a reporting occasion.

[0235] In some aspects, the assistance information is received by the UE or UE provider periodically, or semi-persistently, or in response to a request sent to a network entity.

[0236] In some aspects, the performance report indicates at least one of: an indication of a difference between predicted and actual performance, general system performance, a statistic for each individual inference made by the ML model, or a statistic for each plurality of inferences made by the ML model.

[0237] In some aspects, the performance report indicates the likelihood of fitting in different ML models.

[0238] In some aspects, engaging in changes to the ML model based on the performance report includes engaging in retraining of the ML model or switching to a different ML model.

[0239] In some aspects, the method 1700 further includes: sending an instruction to the network entity to retrain the ML model or switch to a different ML model. In some cases, the operation of this step refers to the operation of Fig.18 The described circuits for transmitting and / or codes for transmitting may be or may be performed by circuits for transmitting and / or codes for transmitting.

[0240] In some aspects, where the indication is to retrain the ML model, the indication also includes information for retraining.

[0241] In some aspects, where the indication is to switch to a different ML model, the indication also includes an identification of the different ML model.

[0242] In some aspects, the indication includes deactivating the ML model.

[0243] In some aspects, the method 1700 further includes: sending an indication for the UE to retrain the ML model or switch to a different ML model. In some cases, the operation of this step refers to the operation of Fig.18 The described circuits for transmitting and / or codes for transmitting may be or may be performed by circuits for transmitting and / or codes for transmitting.

[0244] In some aspects, the indication includes deactivating the ML model.

[0245] In one aspect, method 1700 or any aspect related thereto may be performed by an apparatus such as Fig.18 The method 1700 is performed by a communication device 1800 that includes various components operable, configured, or adapted to perform the method 1700. The communication device 1800 is described in more detail below.

[0246] It should be noted that Fig.17 This is merely one example of a method, and other methods including fewer, additional, or alternative steps may also be consistent with the present disclosure.

[0247] Example Communication Device

[0248] Fig.18 Aspects of an example communication device 1800 are depicted. In some aspects, the communication device 1800 is user equipment, such as described above with respect to Figure 1 and Figure 3 UE 104 is described. In some aspects, the communication device 1800 is a network entity such as Figure 1 and Figure 3 BS 102 or as relative to Figure 2 The decomposed base station in question.

[0249] The communication device 1800 includes a processing system 1805 coupled to a transceiver 1885 (e.g., a transmitter and / or a receiver). In some aspects (e.g., when the communication device 1800 is a network entity), the processing system 1805 can be coupled to a network interface 1895 that is configured to communicate with the communication device 1800 via a communication link (such as, for example, as described herein with respect to Figure 2 The processing system 1805 may be configured to perform processing functions of the communication device 1800, including processing signals received and / or to be transmitted by the communication device 1800.

[0250] The processing system 1805 includes one or more processors 1810. In various aspects, the one or more processors 1810 may represent one or more of the receive processor 358, the transmit processor 364, the TX MIMO processor 366, and / or the controller / processor 380, as shown in FIG. Figure 3 In various aspects, the one or more processors 1810 may represent one or more of the receive processor 338, the transmit processor 320, the TX MIMO processor 330, and / or the controller / processor 340, as described with respect to Figure 3The one or more processors 1810 are coupled to the computer readable medium / memory 1845 via the bus 1880. In some aspects, the computer readable medium / memory 1845 is configured to store instructions (e.g., computer executable code) that, when executed by the one or more processors 1810, cause the one or more processors 1810 to perform operations related to Fig.14 Method 1400 described herein or any aspect thereof; Fig.15 Method 1500 described herein or any aspect thereof; Fig.16 Method 1600 described herein or any aspect thereof; and / or Fig.17 The method 1700 described herein or any aspect related thereto. Note that references to a processor performing a function of the communication device 1800 may include one or more processors 1810 performing that function of the communication device 1800 .

[0251] In the depicted example, computer readable medium / memory 1845 stores code (e.g., executable instructions), such as code 1850 for obtaining, code 1855 for participating, code 1860 for generating, code 1865 for sending, code 1870 for forwarding, and code 1875 for receiving. The processing of code 1850 for obtaining, code 1855 for participating, code 1860 for generating, code 1865 for sending, code 1870 for forwarding, and code 1875 for receiving may cause communication device 1800 to perform: Fig.14 Method 1400 described herein or any aspect thereof; Fig.15 Method 1500 described herein or any aspect thereof; Fig.16 Method 1600 described herein or any aspect thereof; and / or Fig.17 Method 1700 or any aspect related thereto as described.

[0252] The one or more processors 1810 include circuits configured to implement (e.g., execute) code stored in the computer-readable medium / memory 1845, including circuits such as circuits for obtaining 1815, circuits for participating 1820, circuits for generating 1825, circuits for sending 1830, circuits for forwarding 1835, and circuits for receiving 1840. The processing performed by the circuits for obtaining 1815, participating 1820, generating 1825, sending 1830, forwarding 1835, and receiving 1840 may cause the communication device 1800 to perform: Fig.14 Method 1400 described herein or any aspect thereof; Fig.15Method 1500 described herein or any aspect thereof; Fig.16 Method 1600 described herein or any aspect thereof; and / or Fig.17 Method 1700 or any aspect related thereto as described.

[0253] The various components of the communication device 1800 may provide means for performing the following: Fig.14 Method 1400 described herein or any aspect thereof; Fig.15 Method 1500 described herein or any aspect thereof; Fig.16 Method 1600 described herein or any aspect thereof; and / or Fig.17 The method 1700 described herein or any aspect related thereto. For example, a component for sending, transmitting, or outputting for sending may include Figure 3 The transceiver 354 and / or antenna 352 of the UE 104 are shown. Figure 3 The transceiver 332 and / or antenna 334 of the BS 102 are shown, and / or Fig.18 The communication device 1800 is shown with a transceiver 1885 and an antenna 1890. The means for receiving or obtaining may include Figure 3 The transceiver 354 and / or antenna 352 of the UE 104 are shown. Figure 3 The transceiver 332 and / or antenna 334 of BS 102, and / or Fig.18 The communication device 1800 is shown with a transceiver 1885 and an antenna 1890 .

[0254] Sample Clauses

[0255] Specific implementation examples are described in the following numbered clauses:

[0256] Clause 1: A method of wireless communications by a UE, the method comprising: obtaining from a network entity a performance report for an ML model running on at least one of the UE or the network entity; and engaging in changes to the ML model based on the performance report.

[0257] Clause 2: The method of clause 1, further comprising: forwarding the performance report to an entity associated with the UE.

[0258] Clause 3: The method according to any one of clauses 1 and 2, further comprising: sending assistance information to the network entity to assist the network entity in generating the performance report.

[0259] Clause 4: A method according to clause 3, wherein the auxiliary information includes at least one of the following: ground truth information collected by the UE, an ID of the ML model, a corresponding CSI report configuration, and information about a triggering timing or a reporting timing of the CSI report configuration.

[0260] Clause 5: The method of clause 3, wherein the assistance information is sent in response to a request from the network entity, or periodically, or semi-persistently.

[0261] Clause 6: A method according to any one of clauses 1 to 5, wherein the performance report indicates at least one of: an indication of the difference between predicted performance and actual performance, general system performance, a statistic for each individual inference made by the ML model, or a statistic for each multiple inference made by the ML model.

[0262] Clause 7: A method according to any of clauses 1 to 6, wherein the performance report indicates the possibility of adaptation in different ML models.

[0263] Clause 8: A method according to any one of clauses 1 to 7, wherein said participating in the change to the ML model based on the performance report includes participating in retraining of the ML model or switching to a different ML model.

[0264] Clause 9: The method of clause 8, further comprising: receiving an instruction from the network entity to retrain the ML model or switch to the different ML model.

[0265] Clause 10: The method of clause 9, wherein the indication comprises deactivating the ML model.

[0266] Clause 11: The method of clause 9, further comprising: forwarding the indication to an entity associated with the UE.

[0267] Clause 12: The method of clause 9, wherein, in the case where the instruction is to retrain the ML model, the instruction further comprises information for the retraining.

[0268] Clause 13: The method of clause 9, wherein, if the indication is to switch to the different ML model, the indication further comprises an identification of the different ML model.

[0269] Clause 14: The method of clause 9, further comprising sending an acknowledgement of receipt of the indication to the network entity.

[0270] Clause 15: The method of clause 8, further comprising: sending an instruction to the network entity to retrain the ML model or switch to the different ML model.

[0271] Clause 16: The method of clause 15, wherein the instructing comprises deactivating the ML model.

[0272] Clause 17: The method of clause 15, further comprising: receiving the indication from the entity associated with the UE to retrain the current ML model or switch to the different ML model before sending the indication to the network entity.

[0273] Clause 18: The method of clause 15, wherein, if the instruction is to retrain the ML model, the instruction further comprises information for the retraining.

[0274] Clause 19: The method of clause 15, wherein, if the indication is to switch to the different ML model, the indication further comprises an identification of the different ML model.

[0275] Clause 20: A method of wireless communications by a UE, comprising: generating a performance report for a ML model running on at least one of the UE or a network entity; and engaging in changes to the ML model based on the performance report.

[0276] Clause 21: The method of clause 20, further comprising: forwarding the performance report to a network entity.

[0277] Clause 22: The method of any of clauses 20 to 21, further comprising: receiving assistance information from the network entity to assist in generating the performance report.

[0278] Clause 23: A method according to clause 22, wherein the auxiliary information includes at least one of the following: ground truth information collected by the UE, an ID of the ML model, a corresponding CSI report configuration, and information about a triggering timing or a reporting timing of the CSI report configuration.

[0279] Clause 24: A method as set forth in Clause 22, wherein the assistance information is received by the UE or UE vendor periodically, or semi-persistently, or in response to a request sent to the network entity.

[0280] Clause 25: A method according to any one of clauses 20 to 24, wherein the performance report indicates at least one of: an indication of the difference between predicted performance and actual performance, general system performance, a statistic for each individual inference made by the ML model, or a statistic for each multiple inference made by the ML model.

[0281] Clause 26: A method according to any of clauses 20 to 25, wherein the performance report indicates the likelihood of adaptation in different ML models.

[0282] Clause 27: A method according to any one of clauses 20 to 26, wherein said participating in said changes to the ML model based on the performance report comprises participating in retraining of the ML model or switching to a different ML model.

[0283] Clause 28: The method of clause 27, further comprising: sending an instruction to the network entity to retrain the ML model or switch to the different ML model.

[0284] Clause 29: The method of clause 28, further comprising: receiving the indication to retrain the ML model or switch to the different ML model from the entity associated with the UE before sending the indication to the network entity.

[0285] Clause 30: The method of clause 28, wherein, in the case where the instruction is to retrain the ML model, the instruction further comprises information for the retraining.

[0286] Clause 31: The method of clause 28, wherein, if the indication is to switch to the different ML model, the indication further comprises an identification of the different ML model.

[0287] Clause 32: The method of clause 31, wherein the indication comprises deactivating the ML model.

[0288] Clause 33: The method of clause 27, further comprising: receiving an instruction from the network entity to retrain the ML model or switch to the different ML model.

[0289] Clause 34: The method of clause 33, wherein the indication comprises deactivating the ML model.

[0290] Clause 35: A method of wireless communication by a network entity, the method comprising: sending a performance report for an ML model running on at least one of a UE or the network entity; and engaging in changes to the ML model based on the performance report.

[0291] Clause 36: A method as described in clause 35, the method further comprising: receiving assistance information generated by the UE; and using the assistance information when generating the performance report.

[0292] Clause 37: A method according to clause 36, wherein the auxiliary information includes at least one of the following: ground truth information collected by the UE, the ID of the ML model, the corresponding CSI report configuration, and information about the CSI report configuration triggering timing or reporting timing.

[0293] Clause 38: The method of clause 37, wherein the assistance information is received in response to a request from the network entity, or periodically, or semi-persistently.

[0294] Clause 39: A method according to any one of clauses 35 to 38, wherein the performance report indicates at least one of: an indication of the difference between predicted performance and actual performance, general system performance, a statistic for each individual inference made by the ML model, or a statistic for each multiple inference made by the ML model.

[0295] Clause 40: A method according to any of clauses 35 to 39, wherein the performance report indicates the likelihood of adaptation in different ML models.

[0296] Clause 41: A method as described in any of clauses 35 to 40, wherein said engaging in the change to the ML model based on the performance report comprises sending an indication for the UE to retrain the ML model or switch to the different ML model.

[0297] Clause 42: The method of clause 41, wherein the indication comprises deactivating the ML model.

[0298] Clause 43: The method of clause 41, wherein, in the case where the instruction is to retrain the ML model, the instruction further comprises information for the retraining.

[0299] Clause 44: The method of clause 41, wherein, if the indication is to switch to the different ML model, the indication further comprises an identification of the different ML model.

[0300] Clause 45: The method of clause 41, further comprising: receiving an acknowledgement of receipt of the indication by the UE.

[0301] Clause 46: The method of clause 45, further comprising: receiving an indication from the UE to retrain the ML model or switch to the different ML model.

[0302] Clause 47: The method of clause 46, wherein the indication comprises deactivating the ML model.

[0303] Clause 48: A method of wireless communication by a network entity, the method comprising: receiving a performance report for an ML model running on at least one of a UE or the network entity; and engaging in changes to the ML model based on the performance report.

[0304] Clause 49: The method according to clause 48, further comprising: sending assistance information to assist the UE in generating the performance report.

[0305] Clause 50: A method according to clause 49, wherein the auxiliary information includes at least one of the following: ground truth information collected by the UE, an ID of the ML model, a corresponding CSI report configuration, and information about a CSI report configuration triggering timing or a reporting timing.

[0306] Clause 51: The method of clause 49, wherein the assistance information is received by the UE or UE vendor periodically, or semi-persistently, or in response to a request sent to the network entity.

[0307] Clause 52: A method according to any one of clauses 48 to 51, wherein the performance report indicates at least one of: an indication of the difference between predicted performance and actual performance, general system performance, a statistic for each individual inference made by the ML model, or a statistic for each multiple inference made by the ML model.

[0308] Clause 53: The method of clause 52, wherein the performance report indicates the likelihood of adaptation in different ML models.

[0309] Clause 54: The method of clause 52, wherein said engaging in said changes to said ML model based on said performance report comprises engaging in retraining of said ML model or switching to a different ML model.

[0310] Clause 55: The method of clause 54, further comprising: sending an instruction to the network entity to retrain the ML model or switch to the different ML model.

[0311] Clause 56: A method according to clause 55, wherein, if the instruction is to retrain the ML model, the instruction also includes information for the retraining.

[0312] Clause 57: The method of clause 55, wherein, if the indication is to switch to the different ML model, the indication further comprises an identification of the different ML model.

[0313] Clause 58: The method of clause 57, wherein the indication comprises deactivating the ML model.

[0314] Clause 59: A method according to any one of clauses 48 to 58, the method further comprising: sending an indication for the UE to retrain the ML model or switch to the different ML model.

[0315] Clause 60: The method of clause 59, wherein the indication comprises deactivating the ML model.

[0316] Clause 61: An apparatus, the apparatus comprising: a memory comprising executable instructions; and a processor configured to execute the executable instructions and cause the apparatus to perform the method of any one of clauses 1 to 60.

[0317] Clause 62: An apparatus comprising means for performing the method of any of clauses 1 to 60.

[0318] Clause 63: A non-transitory computer readable medium comprising executable instructions which, when executed by a processor of an apparatus, cause the apparatus to perform the method of any of clauses 1 to 60.

[0319] Clause 64: A computer program product embodied in a computer readable storage medium, the computer readable storage medium comprising code for performing the method according to any one of clauses 1 to 60.

[0320] Additional considerations

[0321] The foregoing description is provided to enable any person skilled in the art to practice the various aspects described herein. The examples discussed herein do not limit the scope, applicability, or aspects set forth in the claims. Various modifications to these aspects will be apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects. For example, without departing from the scope of the present disclosure, the functions and arrangements of the elements discussed may be changed. Various examples may omit, replace, or add various procedures or components as appropriate. For example, the described methods may be performed in a different order than the described order, and various actions may be added, omitted, or combined. In addition, the features described for some examples may be combined in some other examples. For example, any number of aspects set forth herein may be used to implement a device or practice method. In addition, the scope of the present disclosure is intended to cover such devices or methods practiced using other structures, functionality, or structures and functionality other than the various aspects of the present disclosure set forth herein or different from the various aspects of the present disclosure set forth herein. It should be understood that any aspect of the present disclosure disclosed herein may be embodied by one or more elements of the claims.

[0322] The various illustrative logical blocks, modules, and circuits described in conjunction with the present disclosure may be implemented or performed using a general purpose processor, a digital signal processor (DSP), an ASIC, a field programmable gate array (FPGA) or other programmable logic device (PLD), 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 in an alternative, the processor may be any commercially available 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, one or more microprocessors combined with a DSP core, a system on a chip (SoC), or any other such configuration.

[0323] As used herein, a phrase referring to "at least one of" a list of items refers to any combination of those items (including single members). As an example, "at least one of a, b, or c" is intended to cover: a, b, c, ab, ac, bc, and abc, as well as any combination with multiple identical elements (e.g., aa, aaa, aab, aac, abb, acc, bb, bbb, bbc, cc, and ccc, or any other ordering of a, b, and c).

[0324] As used herein, the term "determining" encompasses a wide variety of actions. For example, "determining" may include calculating, computing, processing, deriving, investigating, searching (e.g., searching in a table, a database, or another data structure), ascertaining, and the like. Additionally, "determining" may include receiving (e.g., receiving information), accessing (e.g., accessing data in a memory), and the like. Additionally, "determining" may include resolving, selecting, choosing, establishing, and the like.

[0325] The method disclosed herein includes one or more actions for implementing the method. The method actions can be interchangeable with each other without departing from the scope of the claims. In other words, unless a specific order of actions is specified, the order and / or use of specific actions can be modified without departing from the scope of the claims. In addition, the various operations of the method described above can be performed by any suitable component that can perform the corresponding function. The component may include various hardware and / or software components and / or modules, including but not limited to circuits, application specific integrated circuits (ASICs) or processors.

[0326] The following claims are not intended to be limited to the various aspects shown herein, but should be given the full scope consistent with the language of the claims. Within the claims, unless explicitly stated, reference to an element in the singular form is not intended to mean "one and only one", but "one or more". Unless otherwise specified, the term "some" refers to one or more. Any claim element is not interpreted according to the provisions of 35 U.S.C. § 112 (f) unless the element is explicitly stated using the phrase "member for...". All structural and functional equivalents of the elements throughout the various aspects described in this disclosure that are known or will be known later to a person of ordinary skill in the art are expressly incorporated herein by reference and are intended to be covered by the claims. In addition, nothing disclosed herein is intended to be dedicated to the public, regardless of whether such disclosure is explicitly stated in the claims.

Claims

1. A method for wireless communication by a user equipment (UE), the method comprising: obtaining, from a network entity, a performance report for a machine learning (ML) model running on at least one of the UE or the network entity; as well as Involve changes to the ML model based on the performance report.

2. The method according to claim 1, further comprising: The performance report is forwarded to an entity associated with the UE.

3. The method according to claim 1, further comprising: Assistance information is sent to the network entity to assist the network entity in generating the performance report.

4. The method of claim 3, wherein the auxiliary information comprises at least one of the following: ground truth information collected by the UE, an ID of the ML model, a corresponding channel station information (CSI) report configuration, information about a CSI report configuration triggering timing or a reporting timing, or a baseline CSI report using a non-ML-based precoder matrix indicator (PMI) codebook, wherein the baseline CSI report is based on a CSI reference signal (CSI-RS) resource for channel measurement that is the same as an ML-based CSI report.

5. The method of claim 3, wherein the assistance information is sent in response to a request from the network entity, or periodically, or semi-persistently.

6. The method of claim 1 , wherein the performance report indicates at least one of: an indication of a difference between predicted and actual performance, general system performance, a statistic for each individual inference made by the ML model, or a statistic for each plurality of inferences made by the ML model.

7. The method of claim 1, wherein the performance report indicates the likelihood of fitting in different ML models.

8. The method of claim 1, wherein engaging in changes to the ML model based on the performance report comprises engaging in retraining of the ML model or switching to a different ML model.

9. The method according to claim 8, further comprising: An instruction to retrain the ML model or to switch to the different ML model is received from the network entity.

10. The method of claim 9, wherein the instructing comprises deactivating the ML model.

11. The method according to claim 9, further comprising: The indication is forwarded to an entity associated with the UE.

12. The method of claim 9, wherein, in the case where the instruction is to retrain the ML model, the instruction further includes information for the retraining.

13. The method of claim 9, wherein if the instruction is to switch to the different ML model, the instruction further includes an identification of the different ML model.

14. The method according to claim 9, further comprising: An acknowledgement of receipt of the indication is sent to the network entity.

15. The method according to claim 8, further comprising: An instruction to retrain the ML model or switch to the different ML model is sent to the network entity. The method of claim 15 , wherein the instructing comprises deactivating the ML model.

17. The method according to claim 15, further comprising: Prior to sending the indication to the network entity, the indication to retrain the current ML model or to switch to the different ML model is received from the entity associated with the UE.

18. The method of claim 15, wherein in the case where the instruction is to retrain the ML model, the instruction further includes information for the retraining.

19. The method of claim 15, wherein if the instruction is to switch to the different ML model, the instruction further includes an identification of the different ML model.

20. A method for wireless communication by a user equipment (UE), the method comprising: generating a performance report for a machine learning (ML) model running on at least one of the UE or a network entity; as well as Involve changes to the ML model based on the performance report.

21. The method according to claim 20, further comprising: The performance report is forwarded to a network entity.

22. The method according to claim 20, further comprising: Assistance information is received from the network entity to assist in generating the performance report.

23. The method of claim 22, wherein the auxiliary information comprises at least one of the following: ground truth information collected by the UE, an ID of the ML model, a corresponding channel station information (CSI) report configuration, and information about a CSI report configuration triggering timing or a reporting timing.

24. The method of claim 22, wherein the assistance information is received by the UE or UE vendor periodically, or semi-persistently, or in response to a request sent to the network entity.

25. The method of claim 20, wherein the performance report indicates at least one of: an indication of a difference between predicted and actual performance, general system performance, a statistic for each individual inference made by the ML model, or a statistic for each plurality of inferences made by the ML model.

26. The method of claim 20, wherein the performance report indicates the likelihood of fitting in different ML models.

27. The method of claim 20, wherein engaging in changes to the ML model based on the performance report comprises engaging in retraining of the ML model or switching to a different ML model.

28. The method according to claim 27, further comprising: An instruction to retrain the ML model or switch to the different ML model is sent to the network entity.

29. The method according to claim 28, further comprising: Prior to sending the indication to the network entity, the indication to retrain the ML model or switch to the different ML model is received from the entity associated with the UE.

30. The method of claim 28, wherein in the case where the instruction is to retrain the ML model, the instruction further includes information for the retraining.

31. The method of claim 28, wherein if the instruction is to switch to the different ML model, the instruction further includes an identification of the different ML model.

32. The method of claim 31 , wherein the instructing comprises deactivating the ML model.

33. The method of claim 27, further comprising: An instruction to retrain the ML model or to switch to the different ML model is received from the network entity.

34. The method of claim 33, wherein the instructing comprises deactivating the ML model.

35. A method for wireless communication by a network entity, the method comprising: sending a performance report for a machine learning (ML) model running on at least one of a user equipment (UE) or the network entity; as well as Involve changes to the ML model based on the performance report.

36. The method of claim 35, further comprising: receiving assistance information generated by the UE; as well as The auxiliary information is used when generating the performance report.

37. The method of claim 36, wherein the auxiliary information comprises at least one of: ground truth information collected by the UE, an ID of the ML model, a corresponding channel station information (CSI) report configuration, and information about a CSI report configuration triggering timing or a reporting timing.

38. The method of claim 36, wherein receiving the auxiliary information comprises receiving a baseline CSI report using a non-ML-based precoder matrix indicator (PMI) codebook, wherein the baseline CSI report is based on the same CSI reference signal (CSI-RS_resource) for channel measurement as the ML-based CSI report.

39. The method of claim 38, further comprising: A request or configuration for a baseline CSI report is sent using a non-ML-based PMI codebook, wherein the configuration or request includes configuring the same CSI-RS resources for channel measurement for the ML-based CSI report and the baseline CSI report.

40. The method of claim 37, wherein the assistance information is received in response to a request from the network entity, or periodically, or semi-persistently.

41. The method of claim 35, wherein the performance report indicates at least one of: an indication of a difference between predicted and actual performance, general system performance, a statistic for each individual inference made by the ML model, or a statistic for each multiple inferences made by the ML model.

42. The method of claim 35, wherein the performance report indicates the likelihood of fitting in different ML models.

43. The method of claim 35, wherein engaging in changes to the ML model based on the performance report comprises sending an indication for the UE to retrain the ML model or switch to the different ML model.

44. The method of claim 43, wherein the instructing comprises deactivating the ML model.

45. The method of claim 43, wherein in the case where the instruction is to retrain the ML model, the instruction further includes information for the retraining.

46. ​​The method of claim 43, wherein if the indication is to switch to the different ML model, the indication further comprises an identification of the different ML model.

47. The method of claim 43, further comprising: An acknowledgement is received that the UE received the indication.

48. The method of claim 37, further comprising: An indication is received from the UE to retrain the ML model or switch to the different ML model.

49. The method of claim 48, wherein the instructing comprises deactivating the ML model.

50. A method for wireless communication by a network entity, the method comprising: receiving a performance report for a machine learning (ML) model running on at least one of a user equipment (UE) or the network entity; as well as Involve changes to the ML model based on the performance report.

51. The method of claim 50, further comprising: Sending assistance information to assist the UE in generating the performance report.

52. The method of claim 51, wherein the auxiliary information comprises at least one of: ground truth information collected by the UE, an ID of the ML model, a corresponding channel station information (CSI) report configuration, and information about a CSI report configuration triggering timing or a reporting timing.

53. The method of claim 51, wherein the assistance information is received by the UE or UE vendor periodically, or semi-persistently, or in response to a request sent to the network entity.

54. The method of claim 50, wherein the performance report indicates at least one of: an indication of a difference between predicted and actual performance, general system performance, a statistic for each individual inference made by the ML model, or a statistic for each multiple inferences made by the ML model.

55. The method of claim 54, wherein the performance report indicates the likelihood of fitting in different ML models.

56. The method of claim 54, wherein engaging in changes to the ML model based on the performance report comprises engaging in retraining of the ML model or switching to a different ML model.

57. The method of claim 56, further comprising: An instruction to retrain the ML model or switch to the different ML model is sent to the network entity.

58. The method of claim 57, wherein in the case where the instruction is to retrain the ML model, the instruction further comprises information for the retraining.

59. The method of claim 57, wherein if the indication is to switch to the different ML model, the indication further comprises an identification of the different ML model.

60. The method of claim 59, wherein the instructing comprises deactivating the ML model.

61. The method of claim 50, further comprising: Sending an indication for the UE to retrain the ML model or switch to the different ML model.

62. The method of claim 61, wherein the instructing comprises deactivating the ML model.

63. An apparatus, comprising: a memory including executable instructions; and a processor configured to execute the executable instructions and cause the apparatus to perform the method according to any one of claims 1 to 60.

64. An apparatus comprising means for performing the method according to any one of claims 1 to 60.

65. A non-transitory computer readable medium comprising executable instructions which, when executed by a processor of a device, cause the device to perform the method of any one of claims 1 to 60.

66. A computer program product embodied on a computer readable storage medium, the computer readable storage medium comprising code for executing the method according to any one of claims 1 to 60.

Images