Knowledge distillation based multi-vendor split learning techniques for cross-node machine learning

Abstract

The techniques described herein utilize a machine learning algorithm to train encoders from multiple UE providers and shared decoders from gNB providers in order to develop a universal gNB decoder, the universal gNB decoder may be capable of decoding inputs from UEs of different UE providers with performance and overhead comparable to different decoders specially developed for each encoder.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the benefit of U.S. Provisional Application Serial No. 63 / 371,311, filed on August 12, 2022, entitled “TECHNIQUES FOR KNOWLEDGEDISTILLATION BASED MULTI-VENDOR SPLIT LEARNING FOR CROSS-NODE MACHINELEARNING,” and U.S. Application Serial No. 18 / 340,732, filed on June 23, 2023, entitled “TECHNIQUES FOR KNOWLEDGEDISTILLATION BASED MULTI-VENDOR SPLIT LEARNING FOR CROSS-NODE MACHINELEARNING,” both of which are expressly incorporated herein by reference in their entirety. Background Art Technical Field

[0003] The present disclosure relates generally to communication systems, and more particularly to knowledge distillation techniques in multi-vendor split learning for cross-node machine learning (ML).

[0004] introduction

[0005] Wireless communication systems are widely deployed to provide a variety of telecommunication services, such as telephony, video, data, messaging, and broadcast. Typical wireless communication systems may employ multiple access technologies capable of supporting communication with multiple users by sharing available system resources. Examples of such multiple access technologies include code division multiple access (CDMA) systems, time division multiple access (TDMA) systems, frequency division multiple access (FDMA) systems, orthogonal frequency division multiple access (OFDMA) systems, single carrier frequency division multiple access (SC-FDMA) systems, and time division synchronous code division multiple access (TD-SCDMA) systems.

[0006] These multiple access technologies have been adopted in various telecommunication standards to provide a common protocol that enables different wireless devices to communicate at a city, country, region, and even global level. An example telecommunication standard is 5G New Radio (NR). 5GNR is part of the continued mobile broadband evolution promulgated by the Third Generation Partnership Project (3GPP) to meet new requirements associated with latency, reliability, security, scalability (e.g., with the Internet of Things (IoT)), and other requirements. 5G NR includes services associated with enhanced mobile broadband (eMBB), massive machine type communications (mMTC), and ultra-reliable low latency communications (URLLC). Certain aspects of 5G NR may be based on the 4G Long Term Evolution (LTE) standard.

[0007] Therefore, further improvements to 5G NR technology are needed. In addition, these improvements may also be applicable to other multiple access technologies and telecommunication standards that employ these technologies. For example, it is desirable to improve the efficiency and latency associated with the mobility of user equipment (UE) communicating with network entities. Summary of the invention

[0008] The following presents a simplified summary of one or more aspects in order to provide a basic understanding of these aspects. This summary is not an extensive overview of all contemplated aspects, and is neither intended to identify key or important elements of all aspects, nor to describe the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to a more detailed description presented later.

[0009] One example aspect includes a method for training a shared base station (gNB) decoder for wireless communication using a machine learning (ML) algorithm. The method may include encoding a set of channel state information (CSI) pre-decoding vectors via one or more teacher user equipment (UE) encoders. The method may also include decoding the output of the one or more teacher UE encoders through one or more gNB teacher decoders to generate teacher reconstructed CSI vectors. The method may also include calculating a loss function between the teacher reconstructed CSI vectors and a ground truth based on the set of CSI pre-decoding vectors. The method may also include training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the calculations. The method may also include distilling the encoding functionality of the one or more teacher UE encoders into corresponding one or more student UE encoders. The method may also include distilling the decoding functionality of the one or more gNB teacher decoders into the shared gNB decoder, wherein the shared gNB decoder is configured to decode communications received from multiple UEs and multiple wireless network providers.

[0010] Another example aspect includes an apparatus for training a shared base station (gNB) decoder for wireless communication using an ML algorithm, the apparatus comprising one or more memories and one or more processors coupled to the one or more memories. The one or more processors may be configured, individually or in combination, to encode a set of CSI pre-decoding vectors via one or more teacher UE encoders. The one or more processors may also be configured, individually or in combination, to decode the output of the one or more teacher UE encoders through one or more gNB teacher decoders to generate teacher reconstructed CSI vectors. The one or more processors may also be configured, individually or in combination, to calculate a loss function between the teacher reconstructed CSI vectors and a reference truth value based on the set of CSI pre-decoding vectors. The one or more processors may also be configured, individually or in combination, to train the one or more teacher UE encoders and the one or more gNB teacher decoders based on these calculations. The one or more processors may also be configured, individually or in combination, to distill the encoding functionality of the one or more teacher UE encoders into corresponding one or more student UE encoders. The one or more processors may be further configured, individually or in combination, to: distill decoding functionality of the one or more gNB teacher decoders into the shared gNB decoder, wherein the shared gNB decoder is configured to decode communications received from multiple UEs and multiple wireless network providers.

[0011] Another example includes an apparatus for wireless communication by a user equipment, the apparatus including a component for training a shared gNB decoder for wireless communication using an ML algorithm. The apparatus may include a component for encoding a set of CSI pre-decoding vectors via one or more teacher UE encoders. The apparatus may also include a component for decoding the output of the one or more teacher UE encoders through one or more gNB teacher decoders to generate teacher reconstructed CSI vectors. The apparatus may also include a component for calculating a loss function between the teacher reconstructed CSI vectors and a reference truth based on the set of CSI pre-decoding vectors. The apparatus may also include a component for training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the calculations. The apparatus may also include a component for distilling the encoding functionality of the one or more teacher UE encoders into corresponding one or more student UE encoders. The apparatus may also include a component for distilling the decoding functionality of the one or more gNB teacher decoders into the shared gNB decoder, wherein the shared gNB decoder is configured to decode communications received from multiple UEs and multiple wireless network providers.

[0012] Another example includes one or more non-transitory computer-readable media storing instructions, alone or in combination, that are executable by one or more processors for training a shared gNB decoder for wireless communication using an ML algorithm. The instructions executable by the one or more processors include instructions for encoding a set of CSI pre-decoding vectors via one or more teacher UE encoders. The instructions are also executable by the one or more processors for decoding the output of the one or more teacher UE encoders through one or more gNB teacher decoders to generate teacher reconstructed CSI vectors. The instructions are also executable by the one or more processors for calculating a loss function between the teacher reconstructed CSI vectors and a reference truth value based on the set of CSI pre-decoding vectors. The instructions are also executable by the one or more processors for training the one or more teacher UE encoders and the one or more gNB teacher decoders based on these calculations. The instructions are also executable by the one or more processors for distilling the encoding functionality of the one or more teacher UE encoders into corresponding one or more student UE encoders. The instructions are also executable by the one or more processors to distill decoding functionality of the one or more gNB teacher decoders into the shared gNB decoder, wherein the shared gNB decoder is configured to decode communications received from multiple UEs and multiple wireless network providers.

[0013] To achieve the foregoing and related ends, one or more aspects include the features fully described below and particularly pointed out in the claims. The following description and the accompanying drawings set forth in detail some illustrative features of one or more aspects. However, these features are merely indicative of some of the various ways in which the principles of the various aspects may be employed, and this specification is intended to include all such aspects and their equivalents. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1A is a diagram illustrating an example of a wireless communication system and an access network.

[0015] Figure 1B is a diagram illustrating an example of a decomposed base station architecture according to various aspects of the present disclosure.

[0016] Figure 2A is a diagram illustrating an example of a first frame according to various aspects of the present disclosure.

[0017] Figure 2B is a diagram illustrating an example of DL channels within a subframe according to various aspects of the present disclosure.

[0018] Figure 2C is a diagram illustrating an example of a second frame according to various aspects of the present disclosure.

[0019] Figure 2D is a diagram illustrating an example of UL channels within a subframe according to various aspects of the present disclosure.

[0020] Figure 3 is an example of a call flow between multiple UE vendor servers and a gNB vendor in accordance with various aspects of the present disclosure.

[0021] Figure 4 is a diagram of an example multiple UE vendor encoder and shared gNB decoder with a trained network deployed in a wireless communication network in accordance with various aspects of the present disclosure.

[0022] Figure 5A is a diagram of a first step of knowledge distillation based training according to various aspects of the present disclosure, where a teacher encoder and a teacher decoder are trained.

[0023] Figure 5B is a diagram of a second step of knowledge distillation based training according to various aspects of the present disclosure, where a student encoder and a shared decoder are trained.

[0024] Figure 5C is a diagram of subsequent steps of knowledge distillation based training according to various aspects of the present disclosure, where a student encoder and a shared decoder are trained.

[0025] Figure 6 is a schematic diagram of an example implementation of various components of a processing system according to various aspects of the present disclosure.

[0026] Figure 7 is a flow chart of an example of a method of wireless communication implemented by a processing system according to aspects of the present disclosure. DETAILED DESCRIPTION

[0027] In cross-node machine learning (ML), the neural network (NN) can be split into two parts: an encoder on the user equipment (UE) and a decoder on the base station (gNB). The encoder output from the UE can be sent to the gNB as input to the decoder. In one example, the encoder at the UE outputs compressed channel state information (CSI), which can be input to the decoder at the gNB. In turn, the decoder at the gNB can output reconstructed CSI, such as a pre-decoding vector.

[0028] In real-world wireless communication systems, the network includes operating UEs and gNBs from any number of different wireless network providers. For the purposes of this disclosure, "wireless network provider" may be used interchangeably with "wireless network supplier" or "supplier" to refer to the manufacturer and provider of equipment used in a wireless network system, such as base stations and / or modems for UEs that may be implemented by different companies in real-world scenarios. Each wireless network provider may have a unique encoder for the UE and therefore require a unique decoder for the gNB. Multiple decoders at the gNB for processing inputs from different types of encoders may require additional system capabilities that may negatively affect system performance. Therefore, there is a need for a "universal" gNB decoder that may be able to decode inputs from any number of different wireless network provider UEs without sacrificing processing time or speed achieved via otherwise customizable or unique decoders commonly used by different suppliers today.

[0029] However, developing a "universal" gNB decoder or a "shared" gNB decoder that may be common across multiple vendors is challenging. In general, training one shared decoder that achieves optimal performance when paired with each of the UE encoders from multiple UE vendors is challenging, especially when the encoders participating in the training session are heterogeneous in terms of architecture and model complexity. This challenge exists because the lower-power encoder dominates the learning signal of the shared decoder during training, thereby degrading the performance of the higher-power encoder.

[0030] The techniques described herein utilize machine learning algorithms to train encoders and decoders from multiple wireless network providers in order to develop a universal gNB decoder that can decode inputs from different wireless network providers with performance and overhead comparable to different decoders developed specifically for each encoder. Since the operation of cross-node ML-based CSI reporting involves different neural networks from multiple wireless network providers, multiple vendors can participate in ML training to optimize their models together. Because multiple vendors can participate in ML training, the ML training can be referred to as a "multi-vendor training" or "multi-network training" system.

[0031] In the multi-vendor training system described herein, each wireless network provider (e.g., UE vendor and gNB vendor) may utilize its own server that participates in offline training individually. One or more UE vendor servers may communicate with corresponding one or more gNB vendor servers using a server-to-server connection during training. Each UE vendor server may train a UE vendor neural network (NN) (e.g., encoder). Similarly, each gNB vendor server may train its own NN (e.g., decoder). To allow joint training of the encoder and decoder, each UE vendor server may provide a ground truth output for the decoder to each gNB vendor server. The UE vendor server and the gNB vendor server may then exchange gradients and activation values.

[0032] In some examples, each UE vendor server may include a teacher encoder and a student encoder, where the student encoder, once trained, can be deployed to its UE. Similarly, a gNB vendor may have a teacher decoder that can be paired with the teacher encoder of each UE vendor server, and a shared decoder that can be paired with all student encoders. Through this specific implementation and as part of the first step in the process, the technology provided herein discloses one-to-one training of a teacher encoder-decoder pair corresponding to each UE vendor. Each UE vendor server and the gNB vendor server can train their teacher encoder-decoder pairs together until convergence. And as part of step two, the neural network parameters of the teacher encoder and teacher decoder can be frozen, and knowledge distillation-based training of the student encoder and shared decoder can be performed. To achieve this, the loss function may include a regularization term to encourage the student encoder-shared decoder pair to mimic the teacher output. Since the shared decoder plays the role of a student in the knowledge distillation process, the term "student decoder" can be used interchangeably with the term "shared decoder".

[0033] Specifically, each UE vendor server may transmit its output (i.e., activation value) from the last layer of its NN (i.e., encoder) to the gNB vendor server. The gNB vendor server may input the activation value received from each UE vendor server to its NN (i.e., decoder). Each UE vendor server also transmits the ground truth output for the decoder to the gNB vendor server. Then, the loss function may be calculated at the gNB vendor server based on the ground truth output provided by each UE vendor server. The gNB vendor server may back-propagate the gradient to the input of its NN (i.e., decoder). Then, the gradient at the input of the gNB vendor server NN (i.e., decoder) may be transmitted to the UE vendor server. Then, each UE vendor server back-propagates the gradient to the input of its NN (i.e., encoder). By implementing knowledge distillation in multi-vendor split learning for cross-node ML, the techniques described herein and elaborated in more detail below may allow the development of a universal gNB that can be incorporated into a gNB for processing encoded data sent by various UE encoders without requiring a unique decoder for each corresponding UE vendor. Such a system will reduce the hardware cost of the gNB.

[0034] The specific embodiments described below in conjunction with the accompanying drawings are intended as descriptions of various configurations and are not intended to represent the only configurations with which the concepts described herein can be practiced. In order to provide a thorough understanding of the various concepts, the specific embodiments include specific details. However, it is apparent to those skilled in the art that these concepts can be practiced without these specific details. In some instances, well-known structures and components are shown in block diagram form to avoid blurring these concepts.

[0035] Several aspects of telecommunication systems will now be presented with reference to various apparatus and methods. These apparatus and methods will be described in the following detailed description and illustrated in the accompanying drawings by various blocks, components, circuits, processes, algorithms, etc. (collectively referred to as "elements"). These elements may be implemented using electronic hardware, computer software, or any combination thereof. Whether these elements are implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system.

[0036] As an example, an element, or any part of an element, or any combination of elements can be implemented as a "processing system", which includes one or more processors. The example of a processor includes a microprocessor, a microcontroller, a graphics processing unit (GPU), a central processing unit (CPU), an application processor, a digital signal processor (DSP), a reduced instruction set computing (RISC) processor, a system on a chip (SoC), a baseband processor, a field programmable gate array (FPGA), a programmable logic device (PLD), a state machine, a gate logic component, a discrete hardware circuit and other suitable hardware configured to perform various functionalities described throughout the present disclosure. One or more processors in a processing system can execute software. Whether it is referred to as software, firmware, middleware, microcode, hardware description language or other names, software should be widely interpreted as meaning instructions, instruction sets, codes, code segments, program codes, programs, subroutines, software components, applications, software applications, software packages, routines, subroutines, objects, executable files, threads of execution, procedures, functions, etc.

[0037] Therefore, in one or more example embodiments, the described functions can be implemented with hardware, software, or any combination thereof. If implemented with software, the functions can be stored or encoded on a computer-readable medium as one or more instructions or codes. Computer-readable media include computer storage media. Storage media can be any available medium that can be accessed by a computer. By way of example and not limitation, such computer-readable media can include random access memory (RAM), read-only memory (ROM), electrically erasable programmable ROM (EEPROM), optical disk storage devices, magnetic disk storage devices, other magnetic storage devices, a combination of computer-readable media of the above types, or any other medium that can be used to store computer executable code in the form of instructions or data structures that can be accessed by a computer.

[0038] As used herein, a processor, at least one processor, and / or one or more processors configured to perform or operable to perform multiple actions (alone or in combination) is intended to include at least two different processors capable of performing different subsets, overlapping subsets, or non-overlapping subsets of the multiple actions, or a single processor capable of performing all of the multiple actions. In a non-limiting example of multiple processors capable of performing different actions combined in the multiple actions, a description of a processor, at least one processor, and / or one or more processors configured to perform or operable to perform actions X, Y, and Z may include at least a first processor configured to perform or operable to perform a first subset of X, Y, and Z (e.g., perform X) and at least a second processor configured to perform or operable to perform a second subset of X, Y, and Z (e.g., perform Y and Z). Alternatively, the first processor, the second processor, and the third processor may be configured to perform or operable to perform corresponding actions in actions X, Y, and Z, respectively. It should be understood that any combination of one or more processors may each be configured to perform or operable to perform any one of the multiple actions or any combination of the multiple actions.

[0039] As used herein, a memory, at least one memory, and / or one or more memories configured to store or having stored thereon instructions for performing multiple actions that are executable by one or more processors are intended to include at least two different memories capable of storing different subsets, overlapping subsets, or non-overlapping subsets of instructions for performing different subsets, overlapping subsets, or non-overlapping subsets of the multiple actions, or a single memory capable of storing instructions for performing all of the multiple actions. In one non-limiting example of one or more memories (alone or in combination) capable of storing different subsets of instructions for performing different actions of the plurality of actions, a description of a memory, at least one memory, and / or one or more memories configured or operable to store or have stored thereon instructions for performing actions X, Y, and Z may include at least a first memory configured or operable to store or have stored thereon a first subset of instructions for performing a first subset of X, Y, and Z (e.g., instructions to perform X), and at least a second memory configured or operable to store or have stored thereon a second subset of instructions for performing a second subset of X, Y, and Z (e.g., instructions to perform Y and Z). Alternatively, the first memory, the second memory, and the third memory may be configured or have stored thereon a respective one of the first subset of instructions for performing X, the second subset of instructions for performing Y, and the third subset of instructions for performing Z, respectively. It should be understood that any combination of one or more memories may each be configured or operable to store or have stored thereon any instruction or any combination of instructions that can be executed by one or more processors to perform any of a plurality of actions or any combination of the plurality of actions. In addition, one or more processors may each be coupled to at least one of the one or more memories and configured or operable to execute instructions to perform the plurality of actions. For example, in the above non-limiting example of different instruction subsets for performing actions X, Y, and Z, a first processor may be coupled to a first memory storing instructions for performing action X, at least a second processor may be coupled to at least a second memory storing instructions for performing actions Y and Z, and the first processor and the second processor may execute the corresponding instruction subsets in combination to complete the execution of actions X, Y, and Z. Alternatively, three processors may access one of three different memories, each of which stores instructions for performing actions X, Y, or Z, and the three processors may execute the corresponding instruction subsets in combination to complete the execution of actions X, Y, and Z. Alternatively, a single processor may execute instructions stored on a single memory or distributed on multiple memories to complete the execution of actions X, Y, and Z.

[0040] Figure 1Ais a diagram illustrating an example of a wireless communication system 100 (also referred to as a wireless wide area network (WWAN)) that includes a base station 102 (also referred to herein as a network entity), a user equipment (UE) 104, an evolved packet core (EPC) 160, and another core network 190 (e.g., a 5G core (5GC)).

[0041] Base station (or network entity) 102 may include macro cells (high-power cellular base stations) and / or small cells (low-power cellular base stations). Macro cells include base stations. Small cells include femto cells, pico cells, and micro cells. Base station 102 may be configured as a decomposed RAN (D-RAN) or open RAN (O-RAN) architecture, in which functionality is split between multiple units (such as a central unit (CU), one or more distributed units (DUs), or radio units (RUs)). Such an architecture may be configured to utilize a protocol stack that is logically split between one or more units (such as one or more CUs and one or more DUs). In some aspects, a CU may be implemented within an edge RAN node, and in some aspects, one or more DUs may be co-located with a CU, or may be geographically distributed in one or more RAN nodes. A DU may be implemented to communicate with one or more RUs. Any decomposed component in a D-RAN and / or O-RAN architecture may be referred to as a network entity herein.

[0042] The base station 102 configured for 4G Long Term Evolution (LTE) (collectively referred to as Evolved Universal Mobile Telecommunications System (UMTS) Terrestrial Radio Access Network (E-UTRAN)) can interface with the EPC 160 via a first backhaul link 132 (e.g., an S1 interface). The base station 102 configured for 5G New Radio (NR) (collectively referred to as Next Generation RAN (NG-RAN)) can interface with the core network 190 via a second backhaul link 184. Among other functions, the base station 102 can also perform one or more of the following functions: delivery of user data, radio channel encryption and decryption, integrity protection, header compression, mobility control functions (e.g., handover, dual connectivity), inter-cell interference coordination, connection establishment and release, load balancing, distribution of non-access stratum (NAS) messages, NAS node selection, synchronization, radio access network (RAN) sharing, multimedia broadcast multicast service (MBMS), subscriber and equipment tracking, RAN information management (RIM), paging, positioning, and delivery of warning messages. The base stations 102 may communicate with each other directly or indirectly (eg, via the EPC 160 or the core network 190) via the third backhaul link 134 (eg, an X2 interface). The first backhaul link 132, the second backhaul link 184, and the third backhaul link 134 may be wired or wireless.

[0043] Base station 102 can communicate wirelessly with UE 104. Each of base stations 102 can provide communication coverage for a corresponding geographic coverage area 110. There may be overlapping geographic coverage areas 110. For example, a small cell 102' may have a coverage area 110' that overlaps with the coverage area 110 of one or more macro base stations 102. A network including both small cells and macro cells can be referred to as a heterogeneous network. A heterogeneous network can also include a home evolved Node B (eNB) (HeNB), which can provide services to a restricted group called a closed subscriber group (CSG). A communication link 120 between base station 102 and UE 104 can include an uplink (UL) (also referred to as a reverse link) transmission from UE 104 to base station 102 and / or a downlink (DL) (also referred to as a forward link) transmission from base station 102 to UE 104. The communication link 120 can use multiple input multiple output (MIMO) antenna technology, including spatial multiplexing, beamforming and / or transmit diversity. The communication link can be through one or more carriers. Base station 102 / UE 104 may use spectrum with a bandwidth of up to Y MHz (e.g., 5 MHz, 10 MHz, 15 MHz, 20 MHz, 100 MHz, 400 MHz, etc.) for each carrier allocated in the carrier aggregation for a total of up to Yx megahertz (MHz) (x component carriers) for transmission in each direction. The carriers may or may not be adjacent to each other. The allocation of carriers may be asymmetric with respect to DL and UL (e.g., more or fewer carriers may be allocated for DL ​​compared to UL). The component carriers may include a primary component carrier and one or more secondary component carriers. The primary component carrier may be referred to as a primary cell (PCell) and the secondary component carrier may be referred to as a secondary cell (SCell).

[0044] Some UEs 104 may communicate with each other using a device-to-device (D2D) communication link 158. The D2D communication link 158 may use DL / UL WWAN spectrum. 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), and a physical sidelink control channel (PSCCH). D2D communication may be through a variety of wireless D2D communication systems, such as, for example, WiMedia, Bluetooth, ZigBee, Wi-Fi based on the Institute of Electrical and Electronics Engineers (IEEE) 802.11 standard, LTE, or NR.

[0045] The wireless communication system may also include a Wi-Fi access point (AP) 150 that communicates with a Wi-Fi station (STA) 152 via a communication link 154 in, for example, a 5 gigahertz (GHz) unlicensed spectrum. When communicating in the unlicensed spectrum, the STA 152 / AP 150 may perform a clear channel assessment (CCA) to determine whether a channel is available prior to communication.

[0046] The small cell 102' may operate in licensed and / or unlicensed spectrum. When operating in an unlicensed spectrum, the small cell 102' may employ NR and use the same unlicensed spectrum (e.g., 5 GHz, etc.) as used by the Wi-Fi AP 150. The small cell 102' employing NR in the unlicensed spectrum may improve the coverage of the access network and / or increase the capacity of the access network.

[0047] The electromagnetic spectrum is typically subdivided into various categories, bands, channels, etc. based on frequency / wavelength. In 5G NR, two initial operating bands have been identified as frequency range designations FR1 (410 MHz–7.125 GHz) and FR2 (24.25 GHz–52.6 GHz). Frequencies between FR1 and FR2 are typically referred to as mid-band frequencies. Although a portion of FR1 is greater than 6 GHz, FR1 is typically (interchangeably) referred to as the “sub-6 GHz” band in various documents and articles. A similar naming issue sometimes occurs with respect to FR2, which is typically (interchangeably) referred to as the “millimeter wave” band in documents and articles, despite being different from the extremely high frequency (EHF) band (30 GHz–300 GHz) identified as the “millimeter wave” band by the International Telecommunication Union (ITU).

[0048] In view of the above aspects, unless otherwise specifically stated, it should be understood that if the term "below 6 GHz" or the like is used herein, it can broadly represent frequencies that can be less than 6 GHz, can be within FR1, or can include mid-band frequencies. In addition, unless otherwise specifically stated, it should be understood that if the term "millimeter wave" or the like is used herein, it can broadly represent frequencies that can include mid-band frequencies, can be within FR2, or can be within the EHF band.

[0049] The base station 102, whether a small cell 102' or a large cell (e.g., a macro base station), may include and / or be referred to as an eNB, a gNodeB (gNB), or another type of base station. Some base stations, such as gNB 180, may operate in the traditional sub-6 GHz spectrum, in millimeter wave frequencies, and / or near millimeter wave frequencies to communicate with the UE 104. When the gNB 180 operates in millimeter wave or near millimeter wave frequencies, the gNB 180 may be referred to as a millimeter wave base station. The millimeter wave base station 180 may utilize beamforming 182 with the UE 104 to compensate for path loss and short range. The base station 180 and the UE 104 may each include multiple antennas, such as antenna elements, antenna panels, and / or antenna arrays, to facilitate beamforming.

[0050] Base station 180 may send beamformed signals to UE 104 in one or more transmit directions 182'. UE 104 may receive beamformed signals from base station 180 in one or more receive directions 182". UE 104 may also send beamformed signals to base station 180 in one or more transmit directions. Base station 180 may receive beamformed signals from UE 104 in one or more receive directions. Base station 180 / UE 104 may perform beam training to determine the best receive direction and transmit direction for each of base station 180 / UE 104. The transmit direction and receive direction of base station 180 may be the same or different. The transmit direction and receive direction of UE 104 may be the same or different.

[0051] The EPC 160 may include a mobility management entity (MME) 162, other MMEs 164, a serving gateway 166, an MBMS gateway 168, a broadcast multicast service center (BM-SC) 170, and a packet data network (PDN) gateway 172. 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. All user Internet Protocol (IP) packets are delivered 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. The IP services 176 may include the Internet, an intranet, an IP multimedia subsystem (IMS), a PS streaming service, and / or other IP services. The BM-SC 170 may provide functions for MBMS user service configuration and delivery. BM-SC 170 may serve as an entry point for content provider MBMS delivery, may be used to authorize and initiate MBMS bearer services in a public land mobile network (PLMN), and may be used to schedule MBMS delivery. MBMS Gateway 168 may be used to distribute MBMS services to base stations 102 belonging to a multicast broadcast single frequency network (MBSFN) area of ​​a broadcast specific service, and may be responsible for session management (start / stop) and for collecting eMBMS related billing information.

[0052] The core network 190 may include an access and mobility management function (AMF) 192, other AMFs 193, a session management function (SMF) 194, and a user plane function (UPF) 195. The AMF 192 may communicate with a unified data management (UDM) 196. The AMF 192 is a control node that handles signaling between the UE 104 and the core network 190. Typically, the AMF 192 provides quality of service (QoS) flows and session management. All user IP packets are delivered through the UPF 195. The UPF 195 provides UE IP address allocation and other functions. The UPF 195 is connected to an IP service 197. The IP service 197 may include the Internet, an intranet, an IMS, a packet switching (PS) streaming media service, and / or other IP services.

[0053] A base station may include and / or be referred to as a network entity, gNB, Node B, eNB, access point, base transceiver station, radio base station, radio transceiver, transceiver function, basic service set (BSS), extended service set (ESS), transmit receive point (TRP), or some other suitable term. Base station 102 provides an access point to EPC 160 or core network 190 for UE 104. Examples of UE 104 include cellular phones, smart phones, session initiation protocol (SIP) phones, laptops, personal digital assistants (PDAs), satellite radios, global positioning systems, multimedia devices, video devices, digital audio players (e.g., MP3 players), cameras, game consoles, tablets, smart devices, wearable devices, vehicles, electric meters, gas pumps, large or small kitchen appliances, healthcare equipment, implants, sensors / actuators, displays, or any other similarly functional device. Some of UE 104 may be referred to as IoT devices (e.g., parking meters, gas pumps, ovens, vehicles, monitors, cameras, industrial / manufacturing equipment, appliances, vehicles, robots, drones, etc.). IoT UEs may include machine type communication (MTC) / enhanced MTC (eMTC, also known as Category (CAT)-M, Cat M1) UEs, NB-IoT (also known as CATNB1) UEs, and other types of UEs. In the present disclosure, eMTC and NB-IoT may refer to future technologies that may evolve from or be based on these technologies. For example, eMTC may include FeMTC (further eMTC), eFeMTC (enhanced further eMTC), mMTC (massive MTC), etc., and NB-IoT may include eNB-IoT (enhanced NB-IoT), FeNB-IoT (further enhanced NB-IoT), etc. UE 104 may also be referred to as a station, a mobile station, a subscriber station, a mobile unit, a subscriber unit, a wireless unit, a remote unit, a mobile device, a wireless device, a wireless communication device, a remote device, a mobile subscriber station, an access terminal, a mobile terminal, a wireless terminal, a remote terminal, a handset, a user agent, a mobile client, a client, or some other suitable terminology.

[0054] Although the present disclosure may focus on 5G NR, the concepts and aspects described herein may be applicable to other similar areas such as LTE, Advanced LTE (LTE-A), Code Division Multiple Access (CDMA), Global System for Mobile Communications (GSM), and / or other wireless / radio access technologies.

[0055] Figure 1B1 is a diagram illustrating an example of a decomposed base station 101 architecture, any component or element of which may be referred to herein as a network entity. The decomposed base station 101 architecture may include one or more central units (CUs) 103, which may communicate directly with a core network 105 via a backhaul link, or indirectly with the core network 105 through one or more decomposed base station units (such as a near real-time (near RT) RAN intelligent controller (RIC) 107 via an E2 link, or a non-real-time (non-RT) RIC 109 associated with a service management and orchestration (SMO) framework 111, or both). The CU 103 may communicate with one or more distributed units (DUs) 113 via corresponding midhaul links (such as an F1 interface). The DU 113 may communicate with one or more radio units (RUs) 115 via corresponding fronthaul links. The RU 115 may communicate with corresponding UEs 104 via one or more radio frequency (RF) access links. In some implementations, a UE 104 may be served simultaneously by multiple RUs 115.

[0056] Each of these units (e.g., CU 103, DU 113, RU 115, and near-RT RIC 107, non-RT RIC 109, and SMO framework 111) may include or be coupled to one or more interfaces configured to receive or send signals, data, or information (collectively referred to as signals) via a wired or wireless transmission medium. Each of the units or an associated processor or controller that provides instructions to the communication interface of these units may be configured to communicate with one or more of the other units via a transmission medium. For example, these 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, these units may include a wireless interface that may include a receiver, a transmitter, or a transceiver (such as a radio frequency (RF) transceiver) that is configured to receive or send signals, or both, to one or more of the other units on a wireless transmission medium.

[0057] In some aspects, CU 103 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 using an interface that is configured to communicate signals with other control functions hosted by CU 103. CU103 may be configured to handle user plane functionality (i.e., central unit-user plane (CU-UP)), control plane functionality (i.e., central unit-control plane (CU-CP)), or a combination thereof. In some specific implementations, CU 103 may be logically split into one or more CU-UP units and one or more CU-CP units. When implemented in an O-RAN configuration, the CU-UP unit may communicate bidirectionally with the CU-CP unit via an interface (such as an E1 interface). As needed, CU 103 may be implemented to communicate with DU 113 for network control and signaling.

[0058] DU 113 may correspond to a logical unit that includes one or more base station functions for controlling the operation of one or more RUs 115. In some aspects, DU 113 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.) at least in part according to a functional split such as defined by the Third Generation Partnership Project (3GPP). In some aspects, DU 113 may further host one or more low PHY layers. Each layer (or module) may be implemented using an interface that is configured to communicate signals with other layers (and modules) hosted by DU 113 or with control functions hosted by CU 103.

[0059] The lower layer functionality may be implemented by one or more RUs 115. In some deployments, the RU 115 controlled by the DU 113 may correspond to a logical node that hosts RF processing functions or low PHY layer functions (such as performing fast Fourier transform (FFT), inverse FFT (iFFT), digital beamforming, physical random access channel (PRACH) extraction and filtering, etc.), or both, based at least in part on functional splitting (such as lower layer functional splitting). In such an architecture, the RU 115 may be implemented to handle over-the-air (OTA) communications with one or more UEs 104. In some specific implementations, real-time and non-real-time aspects of control plane and user plane communications with the RU 115 may be controlled by the corresponding DU 113. In some scenarios, this configuration may enable the implementation of the DU 113 and the CU 103 in a cloud-based RAN architecture (such as a vRAN architecture).

[0060] The SMO framework 111 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 111 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 111 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 to instantiate 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 103, DU 113, RU 115, and near-RT RIC 107. In some specific implementations, the SMO framework 111 may communicate with hardware aspects of the 4G RAN (such as an open eNB (O-eNB) 117) via the O1 interface. Additionally, in some specific implementations, the SMO framework 111 may communicate directly with one or more RUs 115 via the O1 interface. The SMO framework 111 may also include a non-RT RIC 109 configured to support the functionality of the SMO framework 111 .

[0061] The non-RT RIC 109 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 107. The non-RT RIC 109 may be coupled to or in communication with the near-RT RIC 107 (such as via an A1 interface). The near-RT RIC 107 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 103, one or more DUs 113, or both, and the O-eNB with the near-RT RIC 107.

[0062] In some implementations, in order to generate an AI / ML model to be deployed in the near-RT RIC 107, the non-RT RIC 109 may receive parameters or external enrichment information from an external server. Such information may be utilized by the near-RT RIC 107 and may be received from a non-network data source or from a network function at the SMO framework 111 or the non-RT RIC 109. In some examples, the non-RT RIC 109 or the near-RT RIC 107 may be configured to tune RAN behavior or performance. For example, the non-RT RIC 109 may monitor long-term trends and patterns of performance and employ AI / ML models to perform corrective actions through the SMO framework 111 (such as via reconfiguration of O1) or via the creation of RAN management policies (such as A1 policies).

[0063] FIG. 2A to FIG. 2D is a diagram of various frame structures, resources, and channels used by UE 104 and base station 102 / 180 for communications. Figure 2A FIG200 is a diagram illustrating an example of a first subframe within a 5G NR frame structure. Figure 2B is a diagram 230 illustrating an example of DL channels within a 5G NR subframe. Figure 2C FIG250 is a diagram illustrating an example of a second subframe within a 5G NR frame structure. Figure 2D FIG280 is a diagram illustrating an example of an UL channel within a 5G NR subframe. The 5G NR frame structure may be frequency division duplex (FDD) (wherein for a particular set of subcarriers (carrier system bandwidth), a subframe within the subcarrier set is dedicated to either DL or UL), or may be time division duplex (TDD) (wherein for a particular set of subcarriers (carrier system bandwidth), a subframe within the subcarrier set is dedicated to both DL and UL). Figure 2A , Figure 2C In the example provided, the 5G NR frame structure is assumed to be TDD, where subframe 4 is configured with slot format 28 (most of which are DL), where D is DL, U is UL, and F is flexible between DL / UL, and subframe 3 is configured with slot format 34 (most of which are UL). Although subframes 3 and 4 are shown as having slot formats 34 and 28, respectively, any particular subframe may be configured with any of the various available slot formats 0 to 61. Slot formats 0 and 1 are both DL and both UL, respectively. Other slot formats 2-61 include a mix of DL, UL, and flexible symbols. The slot format is configured for the UE via a received slot format indicator (SFI) (dynamically configured via DL control information (DCI) or semi-statically / statically configured via radio resource control (RRC) signaling). Note that the following description also applies to the 5G NR frame structure as TDD.

[0064] Other wireless communication technologies may have different frame structures and / or different channels. For example, a 10 millisecond (ms) frame may be divided into 10 equally sized subframes (1 ms). Each subframe may include one or more time slots. A subframe may also include a microslot, which may include 7, 4, or 2 symbols. Each time slot may include 7 or 14 symbols, depending on the time slot configuration. For time slot configuration 0, each time slot may include 14 symbols, and for time slot configuration 1, each time slot may include 7 symbols. The symbol on the DL may be a cyclic prefix (CP) orthogonal frequency division multiplexing (OFDM) (CP-OFDM) symbol. The symbol on the UL may be a CP-OFDM symbol (for high throughput scenarios) or a discrete Fourier transform (DFT) spread OFDM (DFT-s-OFDM) symbol (also known as a single carrier frequency division multiple access (SC-FDMA) symbol) (for power-limited scenarios; limited to single-stream transmission). The number of time slots in a subframe is based on the time slot configuration and parameter set. For slot configuration 0, different parameter sets μ0 to 4 allow 1, 2, 4, 8, and 16 slots per subframe, respectively. For slot configuration 1, different parameter sets 0 to 2 allow 2, 4, and 8 slots per subframe, respectively. Accordingly, for slot configuration 0 and parameter set μ, there are 14 symbols per slot and 2 per subframe. μ time slots. The subcarrier spacing and symbol length / duration are functions of the parameter set. The subcarrier spacing can be equal to 2 μ *15 kilohertz (kHz), where μ is parameter set 0 to 4. Therefore, the subcarrier spacing for parameter set μ=0 is 15 kHz, and the subcarrier spacing for parameter set μ=4 is 240 kHz. The symbol length / duration is inversely related to the subcarrier spacing. FIG. 2A to FIG. 2D 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. Within a frame set, there may be one or more different bandwidth parts (BWPs) that are frequency-division multiplexed (see Figure 2B ). Each BWP can have a specific set of parameters.

[0065] A resource grid may be used to represent a frame structure. Each slot includes a resource block (RB) (also referred to as a physical RB (PRB)) extending 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.

[0066] like Figure 2A As illustrated, some of the REs carry reference (pilot) signals (RS) for the UE. The RS may include a demodulation RS (DM-RS) for channel estimation at the UE (indicated as R for a particular configuration). x, where 100x is the port number, but other DM-RS configurations are possible) and a channel state information reference signal (CSI-RS). RS may also include a beam measurement RS (BRS), a beam refinement RS (BRRS), and a phase tracking RS (PT-RS).

[0067] Figure 2B 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 including nine RE groups (REGs), each REG including four consecutive REs in an OFDM symbol. The PDCCH within a BWP may be referred to as a control resource set (CORESET). Additional BWPs may be located at higher and / or lower frequencies on the channel bandwidth. The primary synchronization signal (PSS) may be within symbol 2 of a specific subframe of a frame. The PSS is used by UE 104 to determine subframe / symbol timing and physical layer identification. The secondary synchronization signal (SSS) may be within symbol 4 of a specific subframe of a frame. The SSS is used by the UE to determine the physical layer cell identification group number and radio frame timing. Based on the physical layer identification and the physical layer cell identification group number, the UE may determine the physical cell identifier (PCI). Based on the PCI, the UE may determine the position of the aforementioned DM-RS. 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 (also referred to as an SS block (SSB)). The MIB provides the system bandwidth and the number of RBs in the system frame number (SFN). The physical downlink shared channel (PDSCH) carries user data, broadcast system information not sent over the PBCH such as the system information block (SIB), and paging messages.

[0068] like Figure 2C As illustrated in , some of the REs carry DM-RS for channel estimation at the base station (indicated as R for one specific configuration, but other DM-RS configurations are possible). The UE may send a DM-RS for a physical uplink control channel (PUCCH) and a DM-RS for a physical uplink shared channel (PUSCH). The PUSCH DM-RS may be sent in the first or first two symbols of the PUSCH. Depending on whether a short PUCCH or a long PUCCH is sent and on the specific PUCCH format used, the PUCCH DM-RS may be sent in different configurations. The UE may send a sounding reference signal (SRS). The SRS may be sent 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 teeth of the comb. The SRS may be used by the base station for channel quality estimation to achieve frequency-dependent scheduling of the UL.

[0069] Figure 2DExamples of various UL channels within a subframe of a frame are illustrated. The PUCCH may be located at a position as indicated in one configuration. The PUCCH carries uplink control information (UCI), such as scheduling requests, channel quality indicators (CQI), precoding matrix indicators (PMI), rank indicators (RI), and hybrid automatic repeat request (HARQ) acknowledgement (ACK) / negative acknowledgement (NACK) feedback. The PUSCH carries data and may additionally be used to carry buffer status reports (BSRs), power headroom reports (PHRs), and / or UCI.

[0070] Figure 3 305 and gNB vendor server 310. As described above, in a multi-vendor training system, each wireless network provider (e.g., UE vendor and gNB vendor) may utilize its own server that participates in an offline training session. One or more UE vendor servers may communicate with corresponding one or more gNB vendor servers using a server-to-server connection during training.

[0071] Therefore, if Figure 3 As illustrated, each UE vendor may have a corresponding UE vendor server 305. Each gNB vendor may also have a corresponding gNB vendor server 310. The UE vendor server 305 and the gNB vendor server 310 may participate in the same ML training session to train encoders of multiple UE vendor servers 305 and decoders of gNB vendor servers 310. To allow joint training of encoders and decoders, each UE vendor server may provide ground truth outputs for the decoder to each gNB vendor server. The UE vendor server and the gNB vendor server may then exchange gradients and activation values.

[0072] Specifically, each UE vendor server 305 may transmit its output 315 (i.e., activation values) from the last layer of its NN (i.e., encoder) to the gNB vendor server 310. The gNB vendor server 310 may input the activation values ​​received from each UE vendor server 305 into its NN (i.e., decoder). Each UE vendor server 305 may also transmit the ground truth output 320 for the decoder to the gNB vendor server 310. "Ground truth" in machine learning may refer to information known to be true or real or expected through direct observation and measurement (i.e., empirical evidence), rather than information provided by inference. Therefore, the term "ground truth verification" may refer to the process of collecting appropriate objective (verifiable) data for this test.

[0073] A loss function may then be calculated at the gNB vendor server based on the expected baseline truth output 320 provided by each UE vendor server. A “loss function” measures how much (or how little) the estimate from the ML model differs from its true or expected value. The gNB vendor server 310 may then back-propagate the gradients (e.g., back-propagation of the gradients of the loss function for its NN parameters with respect to the quantization error) to the input of its NN (i.e., decoder). The gradients 325 at the input of the gNB vendor server NN (i.e., decoder) may then be transmitted to the UE vendor server 305. In ML, a “gradient” may be the derivative of a function with more than one input variable. Thus, the gradient measures the change in weights to reduce the error quantized by the loss function. Each UE vendor server 305 then back-propagates the gradients to the input of its NN (i.e., encoder).

[0074] Figure 4 4 is a diagram 400 of an example multiple UE vendor encoder 405 and shared gNB decoder 410 with a trained network deployed in a wireless communication network. The example embodiment illustrates operation when a neural network is deployed with multiple UEs 104 from multiple wireless network UE vendors (e.g., a first group of UEs 104 associated with a first UE vendor and a second group of UEs 104 associated with a second UE vendor) that send CSI feedback to a gNB 102 / 180. The gNB 102 / 180 may be part of a single gNB vendor and include a shared decoder 410 for processing CSI feedback messages from multiple UE encoders 405 associated with multiple UE vendors.

[0075] Therefore, UE i (e.g., first and second UEs 104) may each transmit a potential vector z to gNB 102 / 180 i (Index i=1 represents the first UE and i=2 represents the second UE) The gNB 102 / 180 may utilize a shared decoder (i.e., universal decoder) 410 that may be common to all UEs 104 across multiple vendors to process received information. The development of a “universal” gNB decoder that may be common across multiple vendors and platforms utilizes ML algorithms and knowledge distillation-based training that distills the knowledge learned by the teacher encoder and teacher decoder into the student encoder and shared student decoder.

[0076] Specifically, each wireless network provider (e.g., UE vendor and gNB vendor) may utilize its own servers that participate together in offline training. One or more UE vendor servers may communicate with corresponding one or more gNB vendor servers using a server-to-server connection during training. Each UE vendor server may train a UE vendor neural network (NN) (e.g., encoder). Similarly, each gNB vendor server may train its own NN (e.g., decoder). To allow joint training of the encoder and decoder, each UE vendor server may provide ground truth outputs for the decoder to each gNB vendor server. The UE vendor server and the gNB vendor server may then exchange gradients and activation values.

[0077] In some examples, each UE vendor server may include a teacher encoder and a student encoder, where once the student encoder is trained, the student encoder may be deployed to the UE. Similarly, the gNB vendor server may have a teacher decoder that may be paired with the teacher encoder of each UE vendor server, and a shared decoder that may be paired with all student encoders. Through this specific implementation and as part of the first step in the process, the technology provided herein discloses one-to-one training of teacher encoder-decoder pairs corresponding to each UE vendor. Each UE vendor server and the gNB vendor server may train their teacher encoder-decoder pairs together until convergence. And as part of step two, the neural network parameters of the teacher encoder and teacher decoder may be frozen, and knowledge distillation-based training of the student encoder and shared decoder may be performed. To achieve this, the loss function may include a regularization term to encourage the student encoder-shared decoder pair to mimic the teacher output.

[0078] Figure 5A 500 is a diagram of a first step of knowledge distillation based training, where a teacher encoder and a teacher decoder are trained. In some examples, a first UE vendor and a first gNB vendor may train a corresponding first teacher encoder-decoder pair (e.g., a first teacher encoder 505 and a first teacher decoder 510). Figure 5A Only a single encoder-decoder pair is disclosed, but one of ordinary skill in the art should understand that the same process disclosed herein for training a teacher encoder-decoder pair can be performed for any number of encoder-decoder pairs corresponding to different wireless network providers / vendors.

[0079] The input to the teacher encoder 505 may be CSI which may include a set of ground truth pre-decoding vectors 515. The output 520 of the teacher decoder 510 may be a set of reconstructed pre-decoding vectors from the teacher decoder 510. In some examples, the teacher encoder 505 of the first UE vendor may have the same architecture as the student encoder of the first UE vendor, which may be downloaded to the UE of the first UE vendor by the UE vendor server of the first UE vendor. The teacher decoder 510 of the gNB vendor may have the same architecture as the shared decoder of the gNB vendor, which may be downloaded to the gNB of the gNB vendor by the gNB vendor server of the gNB vendor. For the purpose of training the teacher encoder-decoder, the teacher decoder may adopt an approach that attempts to convert v 1,k 515 (e.g., the input to teacher encoder 505) is in the same direction as 520 (e.g., the output of teacher decoder 510). An example is Where N is the number of CSI vectors. This process can be repeated for any number of encoder-decoder pairs corresponding to different wireless network vendors.

[0080] Figure 5B 525 is a diagram of a second step of knowledge distillation-based training, in which a student encoder and a student decoder are trained. As part of the second step of knowledge distillation-based training, the method may include: Figure 5A ), after training a teacher encoder-decoder pair as part of a wireless network, freezing the neural network parameters of the teacher encoder-decoder pair (e.g., a first teacher encoder 505-a and a first teacher decoder 510-a associated with a first wireless network provider, and a second teacher encoder 505-b and a second teacher decoder 510-b associated with a second wireless network provider).

[0081] The system may then implement student encoders (e.g., a first student encoder 530-a associated with a first wireless network provider and a second student encoder 530-b associated with a second wireless network provider). The shared student decoder 535 may be trained to decode the outputs from the first student encoder 530-a and the second student encoder 530-b and compare the results to the results of the teacher decoder 510. It should be appreciated that while Figure 5B The shared student decoder 535 is illustrated as a separate element, but the shared student decoder 535 may be a single decoder where the input of the shared decoder 535 is switched between the outputs from the first student encoder 530 - a and the second student encoder 530 - b .

[0082] Figure 5C550 is a diagram of subsequent steps of knowledge distillation-based training, in which a student encoder and a student decoder are trained. In some examples, CSI vectors 555 (e.g., first CSI vectors 555-a and second CSI vectors 555-b) may be input into corresponding teacher encoders 505 and student encoders 530. For example, at a first UE vendor server, the first CSI vector 555-a may be input into the first teacher encoder 505-a and the first student encoder 530-a. At a second UE vendor server, the second CSI vector 555-b may be input into the second teacher encoder 505-b and the second student encoder 530-b. The output of the encoder may be transmitted to the gNB vendor server.

[0083] On the gNB vendor server side, the first teacher decoder 510-a may decode the output of the first teacher encoder 505-a received from the first UE vendor server. The student shared decoder 535 may also decode the output of the first student encoder 530-a. The second teacher decoder 510-b may decode the output of the second teacher encoder 505-b received from the second UE vendor server. Moreover, the student shared decoder 535 may also decode the output of the second student encoder 530-b.

[0084] Based on the decoding from each of the teacher decoder 510 and the student shared decoder 535, a loss function 580 may be measured for back propagation. As described above, the loss function may be the sum of the losses for each UE vendor server. The loss function 580 may be calculated using a ground truth value (e.g., the truth value of the input 555 initially input to each of the teacher encoder 505 and the student encoder 530), a student reconstructed CSI 565 for the student encoder of the UE vendor server, and a teacher reconstructed CSI 560 for the teacher encoder. The loss function for each UE vendor server includes a reconstruction loss for each UE vendor (which may be expressed as ), and the knowledge distillation loss for each UE vendor (which can be expressed as ).

[0085] Thus, the reconstruction loss may represent the similarity between the student reconstructed CSI 565 (e.g., the output of the shared decoder 535) and the ground truth parameters 555 (e.g., the original input). The knowledge distillation loss may be calculated based on the similarity between the linearly transformed version of the student reconstructed CSI 565 and the teacher reconstructed CSI 560 for the UE vendor. i The linear transformation performed may be referred to as an "adapter". The adaptor may be implemented by a linear layer.

[0086] Therefore, based on the reconstruction loss and the knowledge distillation loss, the weights and biases of the student encoder and the shared decoder may be continuously adjusted so that the shared decoder output mimics the output of the teacher decoder 510 and the ground truth 555 during the entire ML training session. Moreover, when the ML training of the student encoder 530 and the shared decoder 535 is completed, the server will download the corresponding information to the corresponding devices of the vendors. For example, the server for the gNB vendor may download the shared decoder 535 to the gNB vendor's gNB 102, the first UE vendor server for the first UE vendor may download the first student encoder 530a to the UE of the first UE vendor, the second UE vendor server for the second UE vendor may download the second student encoder 530b to the UE of the second UE vendor, and so on.

[0087] Figure 6 600 is a diagram illustrating an example of a hardware implementation for a processing system 605 to implement a machine learning algorithm to train encoders and decoders from multiple wireless network providers in order to develop a universal gNB decoder that can be capable of decoding inputs from different wireless network providers with performance and overhead comparable to different decoders developed for each encoder. The processing system 614 can be implemented with a bus architecture generally represented by bus 624. The bus 624 can include any number of interconnecting buses and bridges, depending on the specific application of the processing system 614 and the overall design constraints. The bus 624 links together various circuits including one or more processors and / or hardware components (which are represented by processor 604, cross-node ML components 640, and one or more computer-readable media / memories 606). The bus 624 can also link various other circuits such as timing sources, peripherals, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further.

[0088] The processing system 614 may be coupled to the transceiver 610. The transceiver 610 is coupled to one or more antennas 620 to receive and transmit information. The transceiver 610 provides components for communicating with various other devices over a transmission medium. The transceiver 610 receives signals from the one or more antennas 620, extracts information from the received signals, and provides the extracted information to the processing system 614 (specifically, the receiver component 642). The receiver component 642 may receive the application traffic 606 and the optimization request 636. In addition, the transceiver 610 receives information from the processing system 614 (specifically, the transmitter component 644) and generates a signal to be applied to the one or more antennas 620 based on the received information.

[0089] The processing system 614 includes a processor 1104 coupled to one or more computer-readable media / memories 606 (e.g., non-transitory computer-readable media). The processor 604 is responsible for general processing, including executing software stored on the one or more computer-readable media / memories 606. The software, when executed by the processor 604, causes the processing system 614 to perform the various functions described above for any particular device. The one or more computer-readable media / memories 606 can also be used to store data manipulated by the processor 604 when executing the software. The processing system 614 also includes a cross-node ML component 640. The aforementioned components can be software components running in the processor 604, resident / stored in the one or more computer-readable media / memories 606, one or more hardware components coupled to the processor 604, or some combination thereof.

[0090] refer to Figure 7 , an example method 700 for wireless communication according to aspects of the present disclosure may be referred to as Figure 6 The discussed one or more processing systems 605 perform. Although the method 700 is described below with respect to elements of the processing system 605, other components, including servers and network entities, may be used to implement one or more of the steps described herein.

[0091] At block 705, method 700 may include encoding a set of channel state information (CSI) pre-coding vectors via one or more teacher user equipment (UE) encoders. In some examples, the method of block 705 may be performed by processor 604, cross-node ML component 640, and / or one or more other components or subcomponents of the processing system 605. In some implementations, processor 604, cross-node ML component 640, and / or one or more other components or subcomponents of the processing system 605 may be configured to and / or may define means for encoding a set of channel state information (CSI) pre-coding vectors via one or more teacher user equipment (UE) encoders.

[0092] At block 710, method 700 may include decoding the output of the one or more teacher UE encoders through one or more gNB teacher decoders to generate a teacher reconstructed CSI vector. In some examples, the method of block 710 may be performed by processor 604, cross-node ML component 640, and / or one or more other components or subcomponents of the processing system 605. In certain implementations, processor 604, cross-node ML component 640, and / or one or more other components or subcomponents of the processing system 605 may be configured to and / or may define components for decoding the output of the one or more teacher UE encoders through one or more gNB teacher decoders to generate a teacher reconstructed CSI vector.

[0093] At block 715, the method 700 may include calculating a loss function between the teacher reconstructed CSI vectors and a reference truth value based on the set of CSI pre-decoding vectors. In some examples, the method of block 715 may be performed by the processor 604, the cross-node ML component 640, and / or one or more other components or subcomponents of the processing system 605. In some implementations, the processor 604, the cross-node ML component 640, and / or one or more other components or subcomponents of the processing system 605 may be configured to and / or may define a component for calculating a loss function between the teacher reconstructed CSI vectors and a reference truth value based on the set of CSI pre-decoding vectors.

[0094] At block 720, method 700 may include training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the calculations. In some examples, the method of block 720 may be performed by processor 604, cross-node ML component 640, and / or one or more other components or subcomponents of the processing system 605. In some implementations, processor 604, cross-node ML component 640, and / or one or more other components or subcomponents of the processing system 605 may be configured to and / or may define components for training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the calculations.

[0095] At block 725, method 700 may include distilling the encoding functionality of the one or more teacher UE encoders into corresponding one or more student UE encoders. In some examples, the method of block 730 may be performed by processor 604, cross-node ML component 640, and / or one or more other components or subcomponents of the processing system 605. In certain implementations, processor 604, cross-node ML component 640, and / or one or more other components or subcomponents of the processing system 605 may be configured to and / or may define components for distilling the encoding functionality of the one or more teacher UE encoders into corresponding one or more student UE encoders.

[0096] At block 730, the method 700 may include distilling decoding functionality of the one or more gNB teacher decoders to the shared gNB decoder, wherein the shared gNB decoder is configured to decode communications received from multiple UEs and multiple wireless network providers. In some examples, the method of block 730 may be performed by the processor 604, the cross-node ML component 640, and / or one or more other components or subcomponents of the processing system 605. In certain implementations, the processor 604, the cross-node ML component 640, and / or one or more other components or subcomponents of the processing system 605 may be configured to and / or may define means for distilling decoding functionality of the one or more gNB teacher decoders to the shared gNB decoder, wherein the shared gNB decoder is configured to decode communications received from multiple UEs and multiple wireless network providers. In some examples, distilling the decoding functionality to the shared gNB decoder and distilling the encoding functionality to the student UE encoder includes freezing teacher parameters during student training.

[0097] In some examples, distilling the decoding functionality of the one or more gNB teacher decoders to the shared gNB decoder and distilling the encoding functionality of the one or more UE teacher encoders to the corresponding one or more UE student encoders may include: calculating a reconstruction loss between a student reconstructed CSI vector output from the shared gNB decoder and a reference truth value to determine the similarity between the reference truth value and the student reconstructed CSI. The method may also include calculating a knowledge distillation loss based on the similarity between a linearly transformed version of the student reconstructed CSI value and the teacher reconstructed CSI value from the corresponding teacher decoder. In addition, the method may include adjusting one or more parameters of the shared gNB decoder and one or more parameters of the student UE encoder to align the student reconstructed CSI value with the teacher reconstructed CSI value and the reference truth value.

[0098] Some further example clauses

[0099] The following examples are merely illustrative and may be combined with aspects of other embodiments or teachings described herein, but are not limited thereto.

[0100] 1. A method for training a shared base station (gNB) decoder for wireless communication using a machine learning (ML) algorithm, the method comprising:

[0101] encoding, via one or more teacher user equipment (UE) encoders, a set of channel state information (CSI) precoding vectors;

[0102] Decoding the outputs of the one or more teacher UE encoders by one or more gNB teacher decoders to generate teacher reconstructed CSI vectors;

[0103] Calculating a loss function between the teacher reconstructed CSI vector and a reference true value based on the set of CSI pre-decoding vectors;

[0104] training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computing;

[0105] distilling the encoding functionality of the one or more teacher UE encoders into corresponding one or more student UE encoders; and

[0106] Distilling decoding functionality of the one or more gNB teacher decoders into the shared gNB decoder, wherein the shared gNB decoder is configured to decode communications received from multiple UEs and multiple wireless network providers.

[0107] 2. A method as described in clause 1, wherein distilling the decoding functionality to the shared gNB decoder and distilling the encoding functionality to the student UE encoder comprises freezing teacher parameters during student training.

[0108] 3. A method as described in any of the preceding clauses 1 or 2, wherein distilling the decoding functionality of the one or more gNB teacher decoders into the shared gNB decoder and distilling the encoding functionality of the one or more UE teacher encoders into the corresponding one or more UE student encoders comprises:

[0109] A reconstruction loss is calculated between a student reconstructed CSI vector output from the shared gNB decoder and a reference truth value to determine the similarity between the reference truth value and the student reconstructed CSI.

[0110] 4. The method according to clause 3, further comprising:

[0111] The knowledge distillation loss is calculated based on the similarity between the linearly transformed version of the student reconstructed CSI value and the teacher reconstructed CSI value from the corresponding teacher decoder.

[0112] 5. The method according to clause 4, further comprising:

[0113] Adjust one or more parameters of the shared gNB decoder and one or more parameters of the student UE encoder to align the student reconstructed CSI value with the teacher reconstructed CSI value and the reference truth value.

[0114] 6. A method according to any one of the preceding clauses 1 to 5, wherein the teacher reconstructed CSI vector comprises a gradient.

[0115] 7. The method according to clause 6, further comprising:

[0116] Exchange gradients and activation values ​​with the one or more teacher UE encoders.

[0117] 8. The method of clause 6, wherein training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computation further comprises:

[0118] Back-propagate the gradients into the shared gNB decoder.

[0119] 9. An apparatus for training a shared base station (gNB) decoder for wireless communication using a machine learning (ML) algorithm, the apparatus comprising:

[0120] one or more memories; and

[0121] one or more processors, the one or more processors being coupled, individually or in combination, to the one or more memories and configured to:

[0122] encoding, via one or more teacher user equipment (UE) encoders, a set of channel state information (CSI) precoding vectors;

[0123] Decoding the outputs of the one or more teacher UE encoders by one or more gNB teacher decoders to generate teacher reconstructed CSI vectors;

[0124] Calculating a loss function between the teacher reconstructed CSI vector and a reference true value based on the set of CSI pre-decoding vectors;

[0125] training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computing;

[0126] distilling the encoding functionality of the one or more teacher UE encoders into corresponding one or more student UE encoders; and

[0127] Distilling decoding functionality of the one or more gNB teacher decoders into the shared gNB decoder, wherein the shared gNB decoder is configured to decode communications received from multiple UEs and multiple wireless network providers.

[0128] 10. An apparatus as described in clause 9, wherein distilling the decoding functionality to the shared gNB decoder and distilling the encoding functionality to the student UE encoder comprises freezing teacher parameters during student training.

[0129] 11. An apparatus as described in any of the preceding clauses 9 or 10, wherein the one or more processors configured to distill decoding functionality of the one or more gNB teacher decoders to the shared gNB decoder and to distill encoding functionality of the one or more UE teacher encoders to the corresponding one or more UE student encoders are further configured to:

[0130] A reconstruction loss is calculated between a student reconstructed CSI vector output from the shared gNB decoder and a reference truth value to determine the similarity between the reference truth value and the student reconstructed CSI.

[0131] 12. The apparatus of clause 11, wherein the one or more processors are further configured to:

[0132] The knowledge distillation loss is calculated based on the similarity between the linearly transformed version of the student reconstructed CSI value and the teacher reconstructed CSI value from the corresponding teacher decoder.

[0133] 13. The apparatus of clause 12, wherein the one or more processors are further configured to:

[0134] Adjust one or more parameters of the shared gNB decoder and one or more parameters of the student UE encoder to align the student reconstructed CSI value with the teacher reconstructed CSI value and the reference truth value.

[0135] 14. An apparatus according to any of the preceding clauses 9 to 13, wherein the teacher reconstructed CSI vector comprises a gradient.

[0136] 15. The apparatus of clause 14, wherein the one or more processors are further configured to:

[0137] Exchange gradients and activation values ​​with the one or more teacher UE encoders.

[0138] 16. An apparatus as set out in clause 14, wherein the one or more processors configured to train the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computation are further configured to:

[0139] Back-propagate the gradients into the shared gNB decoder.

[0140] 17. One or more non-transitory computer-readable media storing, alone or in combination, instructions executable by one or more processors each coupled to at least one of the one or more non-transitory computer-readable media for training a shared base station (gNB) decoder for wireless communication using a machine learning (ML) algorithm, the one or more non-transitory computer-readable media comprising instructions for:

[0141] encoding, via one or more teacher user equipment (UE) encoders, a set of channel state information (CSI) precoding vectors;

[0142] Decoding the outputs of the one or more teacher UE encoders by one or more gNB teacher decoders to generate teacher reconstructed CSI vectors;

[0143] Calculating a loss function between the teacher reconstructed CSI vector and a reference true value based on the set of CSI pre-decoding vectors;

[0144] training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computing;

[0145] distilling the encoding functionality of the one or more teacher UE encoders into corresponding one or more student UE encoders; and

[0146] Distilling decoding functionality of the one or more gNB teacher decoders into the shared gNB decoder, wherein the shared gNB decoder is configured to decode communications received from multiple UEs and multiple wireless network providers.

[0147] 18. One or more non-transitory computer-readable media as described in clause 17, wherein distilling the decoding functionality to the shared gNB decoder and distilling the encoding functionality to the student UE encoder comprises freezing teacher parameters during student training.

[0148] 19. One or more non-transitory computer-readable media as described in any of the preceding clauses 17 or 18, wherein distilling the decoding functionality of the one or more gNB teacher decoders to the shared gNB decoder and distilling the encoding functionality of the one or more UE teacher encoders to the corresponding one or more UE student encoders comprises:

[0149] A reconstruction loss is calculated between a student reconstructed CSI vector output from the shared gNB decoder and a reference truth value to determine the similarity between the reference truth value and the student reconstructed CSI.

[0150] 20. The one or more non-transitory computer-readable media of clause 19, further comprising instructions for:

[0151] The knowledge distillation loss is calculated based on the similarity between the linearly transformed version of the student reconstructed CSI value and the teacher reconstructed CSI value from the corresponding teacher decoder.

[0152] 21. The one or more non-transitory computer-readable media of clause 20, further comprising instructions for:

[0153] Adjust one or more parameters of the shared gNB decoder and one or more parameters of the student UE encoder to align the student reconstructed CSI value with the teacher reconstructed CSI value and the reference truth value.

[0154] 22. One or more non-transitory computer-readable media according to any one of the preceding clauses 17 to 21, wherein the teacher reconstructed CSI vector comprises a gradient, and wherein the one or more non-transitory computer-readable media further comprises a method for:

[0155] Operation instructions:

[0156] Exchange gradients and activation values ​​with the one or more teacher UE encoders.

[0157] 23. The one or more non-transitory computer-readable media of clause 22, wherein training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computation further comprises:

[0158] Back-propagate the gradients into the shared gNB decoder.

[0159] 24. An apparatus for training a shared base station (gNB) decoder for wireless communication using a machine learning (ML) algorithm, the apparatus comprising:

[0160] means for encoding a set of channel state information (CSI) precoding vectors via one or more teacher user equipment (UE) encoders;

[0161] means for decoding, by one or more gNB teacher decoders, the outputs of the one or more teacher UE encoders to generate a teacher reconstructed CSI vector;

[0162] A component for calculating a loss function between the teacher reconstructed CSI vector and a reference true value based on the set of CSI pre-decoding vectors;

[0163] means for training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computation;

[0164] means for distilling the encoding functionality of the one or more teacher UE encoders into corresponding one or more student UE encoders; and

[0165] Means for distilling decoding functionality of the one or more gNB teacher decoders into the shared gNB decoder, wherein the shared gNB decoder is configured to decode communications received from multiple UEs and multiple wireless network providers.

[0166] 25. An apparatus as described in clause 24, wherein the means for distilling the decoding functionality to the shared gNB decoder and distilling the encoding functionality to the student UE encoder comprises means for freezing teacher parameters during student training.

[0167] 26. An apparatus as described in any of the preceding clauses 24 or 25, wherein the means for distilling the decoding functionality of the one or more gNB teacher decoders to the shared gNB decoder and the means for distilling the encoding functionality of the one or more UE teacher encoders to the corresponding one or more UE student encoders comprises:

[0168] A component for calculating a reconstruction loss between a student reconstructed CSI vector output from the shared gNB decoder and a reference true value to determine a similarity between the reference true value and the student reconstructed CSI.

[0169] 27. The apparatus according to clause 26, further comprising:

[0170] A component for computing the knowledge distillation loss based on the similarity between a linearly transformed version of a student reconstructed CSI value and a teacher reconstructed CSI value from a corresponding teacher decoder.

[0171] 28. The apparatus according to clause 27, further comprising:

[0172] Means for adjusting one or more parameters of the shared gNB decoder and one or more parameters of the student UE encoder to align the student reconstructed CSI value with the teacher reconstructed CSI value and the reference true value.

[0173] 29. An apparatus according to any of the preceding clauses 24 to 28, wherein the teacher reconstructed CSI vector comprises a gradient, and the apparatus further comprises:

[0174] Means for exchanging gradients and activation values ​​with the one or more teacher UE encoders.

[0175] 30. The apparatus of clause 29, wherein the means for training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computation further comprises:

[0176] Means for back-propagating gradients into the shared gNB decoder.

[0177] Although the foregoing disclosure discusses illustrative aspects and / or embodiments, it should be noted that various changes and modifications may be made herein without departing from the described aspects and / or embodiments as defined by the appended claims. Furthermore, although elements of the described aspects and / or embodiments are described or claimed in the singular, the plural is also contemplated unless limitation to the singular is explicitly stated. Additionally, all or part of any aspect and / or embodiment may be used with all or part of any other aspect and / or embodiment unless otherwise stated.

[0178] It should be understood that the specific order or hierarchy of the blocks in the disclosed process / flowchart is merely an illustration of the exemplary method. It should be understood that the specific order or hierarchy of the blocks in the process / flowchart can be rearranged based on design preferences. In addition, some blocks can be combined or omitted. The attached method claims provide the elements of each block in a sample order, but are not meant to be limited to the specific order or hierarchy provided.

[0179] The foregoing description is provided to enable any person skilled in the art to practice the various aspects described herein. Various modifications to these aspects will be apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects. Therefore, the claims are not intended to be limited to the aspects shown herein, but to conform to the full scope consistent with the language claims, wherein the elements mentioned in the singular are not intended to represent "one and only one", unless specifically so stated, but "one or more". Terms such as "if", "when ......" and "while ......" should be interpreted as "under the conditions of ......", rather than meaning an immediate time relationship or reaction. That is, these phrases, such as "when ......", do not mean an immediate action in response to the occurrence of an action or during the occurrence of an action, but simply imply that if the conditions are met, then the action will occur, but no specific or immediate time limit is required for the occurrence of the action. The word "exemplary" is used herein to mean "used as an example, instance, or illustration". Any aspect described as "exemplary" herein is not necessarily interpreted as being preferred or having advantages over other aspects. Unless otherwise specified, the term "some" refers to one or more. Combinations such as “at least one of A, B, or C,” “one or more of A, B, or C,” “at least one of A, B and C,” “one or more of A, B, and C,” and “A, B, C, or any combination thereof” include any combination of A, B, and / or C, and may include multiple A, multiple B, or multiple C. Specifically, combinations such as “at least one of A, B, or C,” “one or more of A, B, or C,” “at least one of A, B, and C,” “one or more of A, B, and C,” and “A, B, C, or any combination thereof” may be only A, only B, only C, A and B, A and C, B and C, or A and B and C, where any such combination may include one or more members of A, B, or C. All structural and functional equivalents to the elements of the various aspects described throughout the disclosure that are known or later become known to those of ordinary skill in the art are expressly incorporated herein by reference and are intended to be encompassed by the claims. Furthermore, nothing disclosed herein is intended to be disclosed to the public, regardless of whether such disclosure is explicitly recited in the claims. Words such as "module," "mechanism," "element," "device," etc. cannot replace the word "component." Therefore, no claim element will be construed as a part-plus-function unless the element is explicitly stated using the phrase "component for..."

Claims

1. A method for training a shared base station (gNB) decoder for wireless communication using a machine learning (ML) algorithm, the method comprising: encoding, via one or more teacher user equipment (UE) encoders, a set of channel state information (CSI) precoding vectors; Decoding the outputs of the one or more teacher UE encoders by one or more gNB teacher decoders to generate teacher reconstructed CSI vectors; Calculating a loss function between the teacher reconstructed CSI vector and a reference true value based on the set of CSI pre-decoding vectors; training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computing; distilling the encoding functionality of the one or more teacher UE encoders into corresponding one or more student UE encoders; as well as Distilling decoding functionality of the one or more gNB teacher decoders into the shared gNB decoder, wherein the shared gNB decoder is configured to decode communications received from multiple UEs and multiple wireless network providers.

2. The method of claim 1 , wherein distilling the decoding functionality to the shared gNB decoder and distilling the encoding functionality to the student UE encoder comprises freezing teacher parameters during student training.

3. The method of claim 1 , wherein distilling the decoding functionality of the one or more gNB teacher decoders into the shared gNB decoder and distilling the encoding functionality of the one or more UE teacher encoders into the corresponding one or more UE student encoders comprises: A reconstruction loss is calculated between a student reconstructed CSI vector output from the shared gNB decoder and a reference truth value to determine the similarity between the reference truth value and the student reconstructed CSI.

4. The method according to claim 3, further comprising: The knowledge distillation loss is calculated based on the similarity between the linearly transformed version of the student reconstructed CSI value and the teacher reconstructed CSI value from the corresponding teacher decoder.

5. The method according to claim 4, further comprising: Adjust one or more parameters of the shared gNB decoder and one or more parameters of the student UE encoder to align the student reconstructed CSI value with the teacher reconstructed CSI value and the reference truth value.

6. The method of claim 1, wherein the teacher reconstructed CSI vector comprises a gradient.

7. The method according to claim 6, further comprising: Exchange gradients and activation values ​​with the one or more teacher UE encoders.

8. The method of claim 6, wherein training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computation further comprises: Back-propagate the gradients into the shared gNB decoder.

9. An apparatus for training a shared base station (gNB) decoder for wireless communication using a machine learning (ML) algorithm, the apparatus comprising: one or more memories; and one or more processors, the one or more processors being coupled, individually or in combination, to the one or more memories and configured to: encoding, via one or more teacher user equipment (UE) encoders, a set of channel state information (CSI) precoding vectors; Decoding the outputs of the one or more teacher UE encoders by one or more gNB teacher decoders to generate teacher reconstructed CSI vectors; Calculating a loss function between the teacher reconstructed CSI vector and a reference true value based on the set of CSI pre-decoding vectors; training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computing; distilling the encoding functionality of the one or more teacher UE encoders into corresponding one or more student UE encoders; as well as Distilling decoding functionality of the one or more gNB teacher decoders into the shared gNB decoder, wherein the shared gNB decoder is configured to decode communications received from multiple UEs and multiple wireless network providers.

10. The apparatus of claim 9, wherein distilling the decoding functionality to the shared gNB decoder and distilling the encoding functionality to the student UE encoder comprises freezing teacher parameters during student training.

11. The apparatus of claim 9, wherein the one or more processors configured to distill decoding functionality of the one or more gNB teacher decoders to the shared gNB decoder and to distill encoding functionality of the one or more UE teacher encoders to the corresponding one or more UE student encoders are further configured to: A reconstruction loss is calculated between a student reconstructed CSI vector output from the shared gNB decoder and a reference truth value to determine the similarity between the reference truth value and the student reconstructed CSI.

12. The apparatus of claim 9, wherein the one or more processors are further configured to: The knowledge distillation loss is calculated based on the similarity between the linearly transformed version of the student reconstructed CSI value and the teacher reconstructed CSI value from the corresponding teacher decoder.

13. The apparatus of claim 12, wherein the one or more processors are further configured to: Adjust one or more parameters of the shared gNB decoder and one or more parameters of the student UE encoder to align the student reconstructed CSI value with the teacher reconstructed CSI value and the reference truth value.

14. The apparatus of claim 9, wherein the teacher reconstructed CSI vector comprises a gradient.

15. The apparatus of claim 14, wherein the one or more processors are further configured to: Exchange gradients and activation values ​​with the one or more teacher UE encoders.

16. The apparatus of claim 14, wherein the one or more processors configured to train the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computation are further configured to: Back-propagate the gradients into the shared gNB decoder.

17. One or more non-transitory computer-readable media storing, alone or in combination, instructions executable by one or more processors each coupled to at least one of the one or more non-transitory computer-readable media for training a shared base station (gNB) decoder for wireless communication using a machine learning (ML) algorithm, the one or more non-transitory computer-readable media comprising instructions for: encoding, via one or more teacher user equipment (UE) encoders, a set of channel state information (CSI) precoding vectors; Decoding the outputs of the one or more teacher UE encoders by one or more gNB teacher decoders to generate teacher reconstructed CSI vectors; Calculating a loss function between the teacher reconstructed CSI vector and a reference true value based on the set of CSI pre-decoding vectors; training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computing; distilling the encoding functionality of the one or more teacher UE encoders into corresponding one or more student UE encoders; as well as Distilling decoding functionality of the one or more gNB teacher decoders into the shared gNB decoder, wherein the shared gNB decoder is configured to decode communications received from multiple UEs and multiple wireless network providers.

18. The one or more non-transitory computer-readable media of claim 17, wherein distilling the decoding functionality to the shared gNB decoder and distilling the encoding functionality to the student UE encoder comprises freezing teacher parameters during student training.

19. The one or more non-transitory computer-readable media of claim 17, wherein distilling decoding functionality of the one or more gNB teacher decoders to the shared gNB decoder and distilling encoding functionality of the one or more UE teacher encoders to the corresponding one or more UE student encoders comprises: A reconstruction loss is calculated between a student reconstructed CSI vector output from the shared gNB decoder and a reference truth value to determine the similarity between the reference truth value and the student reconstructed CSI.

20. The one or more non-transitory computer-readable media of claim 19, further comprising instructions for: The knowledge distillation loss is calculated based on the similarity between the linearly transformed version of the student reconstructed CSI value and the teacher reconstructed CSI value from the corresponding teacher decoder.

21. The one or more non-transitory computer readable media of claim 20, further comprising instructions for: Adjust one or more parameters of the shared gNB decoder and one or more parameters of the student UE encoder to align the student reconstructed CSI value with the teacher reconstructed CSI value and the reference truth value.

22. The one or more non-transitory computer-readable media of claim 17, wherein the teacher reconstructed CSI vector comprises a gradient, and wherein the one or more non-transitory computer-readable media further comprises instructions for: Exchange gradients and activation values ​​with the one or more teacher UE encoders.

23. The one or more non-transitory computer-readable media of claim 22, wherein training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computation further comprises: Back-propagate the gradients into the shared gNB decoder.

24. An apparatus for training a shared base station (gNB) decoder for wireless communication using a machine learning (ML) algorithm, the apparatus comprising: means for encoding a set of channel state information (CSI) precoding vectors via one or more teacher user equipment (UE) encoders; means for decoding, by one or more gNB teacher decoders, the outputs of the one or more teacher UE encoders to generate a teacher reconstructed CSI vector; A component for calculating a loss function between the teacher reconstructed CSI vector and a reference true value based on the set of CSI pre-decoding vectors; means for training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computation; means for distilling the encoding functionality of the one or more teacher UE encoders into corresponding one or more student UE encoders; and Means for distilling decoding functionality of the one or more gNB teacher decoders into the shared gNB decoder, wherein the shared gNB decoder is configured to decode communications received from multiple UEs and multiple wireless network providers.

25. The apparatus of claim 24, wherein the means for distilling the decoding functionality to the shared gNB decoder and the encoding functionality to the student UE encoder comprises means for freezing teacher parameters during student training.

26. The apparatus of claim 24, wherein the means for distilling decoding functionality of the one or more gNB teacher decoders to the shared gNB decoder and the means for distilling encoding functionality of the one or more UE teacher encoders to the corresponding one or more UE student encoders comprises: A component for calculating a reconstruction loss between a student reconstructed CSI vector output from the shared gNB decoder and a reference true value to determine a similarity between the reference true value and the student reconstructed CSI.

27. The apparatus according to claim 26, further comprising: A component for computing the knowledge distillation loss based on the similarity between a linearly transformed version of a student reconstructed CSI value and a teacher reconstructed CSI value from a corresponding teacher decoder.

28. The apparatus according to claim 27, further comprising: Means for adjusting one or more parameters of the shared gNB decoder and one or more parameters of the student UE encoder to align the student reconstructed CSI value with the teacher reconstructed CSI value and the reference true value.

29. The apparatus of claim 24, wherein the teacher reconstructed CSI vector comprises a gradient, and the apparatus further comprises: Means for exchanging gradients and activation values ​​with the one or more teacher UE encoders.

30. The apparatus of claim 29, wherein the means for training the one or more teacher UE encoders and the one or more gNB teacher decoders based on the computation further comprises: Means for back-propagating gradients into the shared gNB decoder.

Images