Apparatus and method for configuring a two-sided model for reporting in a wireless communication system

Abstract

Various aspects of the present disclosure relate to methods, apparatuses, and devices for wireless communication. A first apparatus may determine (802) a first set of one or more parameters for an encoder model of a two-sided model. The first apparatus may determine (804) a first set of information comprising samples that represent an input to the two-sided model. The first apparatus may also update (806) the encoder model according to one or more of a third set of one or more parameters associated with a pre-encoder model, a first message from another apparatus, or a second set of one or more parameters for a pre-encoder generator model, wherein the third set of one or more parameters is based at least in part on the first set of information. The first apparatus may encode (808) data according to the updated encoder model and the first set of information.

Description

APPARATUS AND METHOD FOR CONFIGURING A TWO-SIDED MODEL OF A WIRELESS COMMUNICATION SYSTEM TECHNICAL FIELD

[0001] The present disclosure relates to wireless communications, and more specifically to configuring a two-sided model of a wireless communication system. BACKGROUND

[0002] A wireless communications system may include one or multiple network communication devices, such as base stations, which may support wireless communications for one or multiple user communication devices, which may be otherwise known as user equipment (UE), or other suitable terminology. The wireless communications system may support wireless communications with one or multiple user communication devices by utilizing resources of the wireless communication system (e.g., time resources (e.g., symbols, slots, subframes, frames, or the like) or frequency resources (e.g., subcarriers, carriers, or the like). Additionally, the wireless communications system may support wireless communications across various radio access technologies including third generation (3G) radio access technology, fourth generation (4G) radio access technology, fifth generation (5G) radio access technology, among other suitable radio access technologies beyond 5G (e.g., sixth generation (6G)). SUMMARY

[0003] An article “a” before an element is unrestricted and understood to refer to “at least one” of those elements or “one or more” of those elements. The terms “a,” “at least one,” “one or more,” and “at least one of one or more” may be interchangeable. As used herein, including in the claims, “or” as used in a list of items (e.g., a list of items prefaced by a phrase such as “at least one of” or “one or more of” or “one or both of”) indicates an inclusive list such that, for example, a list of at least one of A, B, or C means A or B or C or AB or AC or BC or ABC (i.e., A and B and C). Also, as used herein, the phrase “based on” shall not be construed as a reference to a closed set of conditions. For example, an example step that is described as “based on condition A” may be based on both a condition A and a condition B without departing from the scope of the present disclosure. In other words, as used herein, the phrase “based on” shall beconstrued in the same manner as the phrase “based at least in part on.” Further, as used herein, including in the claims, a “set” may include one or more elements.

[0004] Various aspects of the present disclosure relate to wireless communications, including improved methods and apparatuses that support configuring a two-sided model (e.g., encoder and decoder) of a wireless communication system. A first apparatus may determine a first set of one or more parameters for an encoder model of a two-sided model. The first apparatus may also determine a first set of information comprising samples that represent an input to the two-sided model. The first apparatus may update the encoder model according to one or more of a third set of one or more parameters associated with a pre-encoder model, a first message from another apparatus, or a second set of one or more parameters for a pre-encoder generator model, wherein the third set of one or more parameters is based at least in part on the first set of information. The first apparatus may also encode data according to the updated encoder model and the first set of information. The first apparatus may transmit the encoded data to a second apparatus. BRIEF DESCRIPTION OF THE DRAWINGS

[0005] Figure 1 illustrates an example of a wireless communications system in accordance with aspects of the present disclosure.

[0006] Figure 2 illustrates an example of a wireless network in accordance with aspects of the present disclosure.

[0007] Figure 3 illustrates an example of a block diagram of a two-sided model in accordance with aspects of the present disclosure.

[0008] Figure 4 illustrates an example of a block diagram of another two-sided model in accordance with aspects of the present disclosure.

[0009] Figure 5 illustrates an example of a UE in accordance with aspects of the present disclosure.

[0010] Figure 6 illustrates an example of a processor in accordance with aspects of the present disclosure.

[0011] Figure 7 illustrates an example of a network equipment (NE) in accordance with aspects of the present disclosure.

[0012] Figure 8 illustrates a flowchart of a method performed by a first apparatus in accordance with aspects of the present disclosure.

[0013] Figure 9 illustrates a flowchart of a method performed by a second apparatus in accordance with aspects of the present disclosure.

[0014] Figure 10 illustrates a flowchart of another method performed by a second apparatus in accordance with aspects of the present disclosure.

[0015] Figure 11 illustrates a flowchart of another method performed by a first apparatus in accordance with aspects of the present disclosure. DETAILED DESCRIPTION

[0016] Various aspects of the present disclosure relate to supporting (e.g., configuring, enabling) a two-sided model of a wireless communication system. The two-sided model may be associated with an encoder of a first wireless device (which may be referred to as an encoder device) and a decoder of the first wireless device or a second wireless device (which may be referred to as a decoder device). Some functions (e.g., operations, behaviors, features) of the model may be performed by the first wireless device (e.g., the encoder device) and other functions (e.g., operations, behaviors, features) of the model may be performed by the first wireless device or the second wireless device (e.g., the decoder device). In some implementations, one or more of the first wireless device (e.g., the encoder device) or the second wireless device (e.g., the decoder device) may manage (e.g., update, adjust, modify) a set of one or more parameters of the two-sided model at a corresponding frequency (e.g., rate, pattern, interval). However, in some cases, excessive data may be used, for example, based on a quantity of data transferred to update the two-sided model and a frequency (e.g., rate) of the updates. By reducing one or more of the transfers including the quantity of data for updating the two-sided model or the frequency (e.g., interval) of updating the two-sided, apparatuses (e.g., wireless devices) performing one or more of encoding or decoding may reduce power consumption, reduce processor usage, reduce data usage, and increase overall system performance.

[0017] Aspects of the present disclosure are described in the context of a wireless communications system.

[0018] Figure 1 illustrates an example of a wireless communications system 100 in accordance with aspects of the present disclosure. The wireless communications system 100 may include one or more NE 102, one or more UE 104, and a core network (CN) 106. The wireless communications system 100 may support various radio access technologies. In some implementations, the wireless communications system 100 may be a 4G network, such as an LTE network or an LTE-Advanced (LTE-A) network. In some other implementations, the wireless communications system 100 may be a new radio (NR) network, such as a 5G network, a 5G-Advanced (5G-A) network, or a 5G ultrawideband (5G-UWB) network. In other implementations, the wireless communications system 100 may be a combination of a 4G network and a 5G network, or other suitable radio access technology including Institute of Electrical and Electronics Engineers (IEEE) 802.11 (Wi-Fi), IEEE 802.16 (WiMAX), IEEE 802.20. The wireless communications system 100 may support radio access technologies beyond 5G, for example, 6G. Additionally, the wireless communications system 100 may support technologies, such as time division multiple access (TDMA), frequency division multiple access (FDMA), or code division multiple access (CDMA), etc.

[0019] The one or more NE 102 may be dispersed throughout a geographic region to form the wireless communications system 100. One or more of the NE 102 described herein may be or include or may be referred to as a network node, a base station, a network element, a network function, a network entity, a radio access network (RAN), a NodeB, an eNodeB (eNB), a next-generation NodeB (gNB), or other suitable terminology. An NE 102 and a UE 104 may communicate via a communication link, which may be a wireless or wired connection. For example, an NE 102 and a UE 104 may perform wireless communication (e.g., receive signaling, transmit signaling) over a Uu interface.

[0020] An NE 102 may provide a geographic coverage area for which the NE 102 may support services for one or more UEs 104 within the geographic coverage area. For example, an NE 102 and a UE 104 may support wireless communication of signals related to services (e.g., voice, video, packet data, messaging, broadcast, etc.) according to one or multiple radio access technologies. In some implementations, an NE 102 may be moveable, for example, a satellite associated with a NTN. In some implementations, different geographic coverage areas associated with the same or different radio accesstechnologies may overlap, but the different geographic coverage areas may be associated with different NE 102.

[0021] The one or more UE 104 may be dispersed throughout a geographic region of the wireless communications system 100. A UE 104 may include or may be referred to as a remote unit, a mobile device, a wireless device, a remote device, a subscriber device, a transmitter device, a receiver device, or some other suitable terminology. In some implementations, the UE 104 may be referred to as a unit, a station, a terminal, or a client, among other examples. Additionally, or alternatively, the UE 104 may be referred to as an Internet-of-Things (IoT) device, an Internet-of-Everything (IoE) device, or machine-type communication (MTC) device, among other examples.

[0022] A UE 104 may be able to support wireless communication directly with other UEs 104 over a communication link. For example, a UE 104 may support wireless communication directly with another UE 104 over a device-to-device (D2D) communication link. In some implementations, such as vehicle-to-vehicle (V2V) deployments, vehicle-to-everything (V2X) deployments, or cellular-V2X deployments, the communication link may be referred to as a sidelink. For example, a UE 104 may support wireless communication directly with another UE 104 over a UE-to-UE interface (PC5 interface).

[0023] An NE 102 may support communications with the CN 106, or with another NE 102, or both. For example, an NE 102 may interface with other NE 102 or the CN 106 through one or more backhaul links (e.g., S1, N2, N2, or network interface). In some implementations, the NE 102 may communicate with each other directly. In some other implementations, the NE 102 may communicate with each other or indirectly (e.g., via the CN 106. In some implementations, one or more NE 102 may include subcomponents, such as an access network entity, which may be an example of an access node controller (ANC). An ANC may communicate with the one or more UEs 104 through one or more other access network transmission entities, which may be referred to as a radio heads, smart radio heads, or transmission-reception points (TRPs).

[0024] The CN 106 may support user authentication, access authorization, tracking, connectivity, and other access, routing, or mobility functions. The CN 106 may be an evolved packet core (EPC), or a 5G core (5GC), which may include a control plane entity that manages access and mobility (e.g., a mobility management entity (MME), anaccess and mobility management functions (AMF)) and a user plane entity that routes packets or interconnects to external networks (e.g., a serving gateway (S-GW), a Packet Data Network (PDN) gateway (P-GW), or a user plane function (UPF)). In some implementations, the control plane entity may manage non-access stratum (NAS) functions, such as mobility, authentication, and bearer management (e.g., data bearers, signal bearers, etc.) for the one or more UEs 104 served by the one or more NE 102 associated with the CN 106.

[0025] The CN 106 may communicate with a packet data network over one or more backhaul links (e.g., via an S1, N2, N2, or another network interface). The packet data network may include an application server. In some implementations, one or more UEs 104 may communicate with the application server. A UE 104 may establish a session (e.g., a protocol data unit (PDU) session, or the like) with the CN 106 via an NE 102. The CN 106 may route traffic (e.g., control information, data, and the like) between the UE 104 and the application server using the established session (e.g., the established PDU session). The PDU session may be an example of a logical connection between the UE 104 and the CN 106 (e.g., one or more network functions of the CN 106).

[0026] In the wireless communications system 100, the NEs 102 and the UEs 104 may use resources of the wireless communications system 100 (e.g., time resources (e.g., symbols, slots, subframes, frames, or the like) or frequency resources (e.g., subcarriers, carriers)) to perform various operations (e.g., wireless communications). In some implementations, the NEs 102 and the UEs 104 may support different resource structures. For example, the NEs 102 and the UEs 104 may support different frame structures. In some implementations, such as in 4G, the NEs 102 and the UEs 104 may support a single frame structure. In some other implementations, such as in 5G and among other suitable radio access technologies, the NEs 102 and the UEs 104 may support various frame structures (i.e., multiple frame structures). The NEs 102 and the UEs 104 may support various frame structures based on one or more numerologies.

[0027] One or more numerologies may be supported in the wireless communications system 100, and a numerology may include a subcarrier spacing and a cyclic prefix. A first numerology (e.g., ^=0) may be associated with a first subcarrier spacing (e.g., 15 kHz) and a normal cyclic prefix. In some implementations, the first numerology (e.g., ^=0) associated with the first subcarrier spacing (e.g., 15 kHz) mayutilize one slot per subframe. A second numerology (e.g., ^=1) may be associated with a second subcarrier spacing (e.g., 30 kHz) and a normal cyclic prefix. A third numerology (e.g., ^=2) may be associated with a third subcarrier spacing (e.g., 60 kHz) and a normal cyclic prefix or an extended cyclic prefix. A fourth numerology (e.g., ^=3) may be associated with a fourth subcarrier spacing (e.g., 120 kHz) and a normal cyclic prefix. A fifth numerology (e.g., ^=4) may be associated with a fifth subcarrier spacing (e.g., 240 kHz) and a normal cyclic prefix.

[0028] A time interval of a resource (e.g., a communication resource) may be organized according to frames (also referred to as radio frames). Each frame may have a duration, for example, a 10 millisecond (ms) duration. In some implementations, each frame may include multiple subframes. For example, each frame may include 10 subframes, and each subframe may have a duration, for example, a 1 ms duration. In some implementations, each frame may have the same duration. In some implementations, each subframe of a frame may have the same duration.

[0029] Additionally, or alternatively, a time interval of a resource (e.g., a communication resource) may be organized according to slots. For example, a subframe may include a number (e.g., quantity) of slots. The number of slots in each subframe may also depend on the one or more numerologies supported in the wireless communications system 100. For instance, the first, second, third, fourth, and fifth numerologies (i.e., ^=0, ^=1, ^=2, ^=3, ^=4) associated with respective subcarrier spacings of 15 kHz, 30 kHz, 60 kHz, 120 kHz, and 240 kHz may utilize a single slot per subframe, two slots per subframe, four slots per subframe, eight slots per subframe, and 16 slots per subframe, respectively. Each slot may include a number (e.g., quantity) of symbols (e.g., orthogonal frequency division multiplexing (OFDM) symbols). In some implementations, the number (e.g., quantity) of slots for a subframe may depend on a numerology. For a normal cyclic prefix, a slot may include 14 symbols. For an extended cyclic prefix (e.g., applicable for 60 kHz subcarrier spacing), a slot may include 12 symbols. The relationship between the number of symbols per slot, the number of slots per subframe, and the number of slots per frame for a normal cyclic prefix and an extended cyclic prefix may depend on a numerology. It should be understood that reference to a first numerology (e.g., ^=0) associated with a first subcarrier spacing (e.g., 15 kHz) may be used interchangeably between subframes and slots.

[0030] In the wireless communications system 100, an electromagnetic (EM) spectrum may be split, based on frequency or wavelength, into various classes, frequency bands, frequency channels, etc. By way of example, the wireless communications system 100 may support one or multiple operating frequency bands, such as frequency range designations FR1 (410 MHz – 7.125 GHz), FR2 (24.25 GHz – 52.6 GHz), FR3 (7.125 GHz – 24.25 GHz), FR4 (52.6 GHz – 114.25 GHz), FR4a or FR4-1 (52.6 GHz – 71 GHz), and FR5 (114.25 GHz – 300 GHz). In some implementations, the NEs 102 and the UEs 104 may perform wireless communications over one or more of the operating frequency bands. In some implementations, FR1 may be used by the NEs 102 and the UEs 104, among other equipment or devices for cellular communications traffic (e.g., control information, data). In some implementations, FR2 may be used by the NEs 102 and the UEs 104, among other equipment or devices for short-range, high data rate capabilities.

[0031] FR1 may be associated with one or multiple numerologies (e.g., at least three numerologies). For example, FR1 may be associated with a first numerology (e.g., ^=0), which includes 15 kHz subcarrier spacing; a second numerology (e.g., ^=1), which includes 30 kHz subcarrier spacing; and a third numerology (e.g., ^=2), which includes 60 kHz subcarrier spacing. FR2 may be associated with one or multiple numerologies (e.g., at least 2 numerologies). For example, FR2 may be associated with a third numerology (e.g., ^=2), which includes 60 kHz subcarrier spacing; and a fourth numerology (e.g., ^=3), which includes 120 kHz subcarrier spacing.

[0032] Figure 2 illustrates an example of a wireless network 200 in accordance with aspects of the present disclosure. The wireless network 200 may include a NE 102-a (e.g., one embodiment of a NE 102, a gNB), a first UE 104-a (e.g., one embodiment of a UE 104, UE1), a second UE 104-b (e.g., one embodiment of a UE 104, UE2), and a third UE 104-c (e.g., one embodiment of a UE 104, UEK, the third UE 104-c represents any number of additional UEs).

[0033] Specifically, Figure 2 shows one example of the wireless network 200 with one NE 102-a (e.g., may be represented by node ^^and may be equipped with ^antennas) and ^ UEs 104-a, 104-b, 104-c (e.g., may be denoted by ^^, ^^ , ⋯ , ^^ thateach have ^ antennas). ^^^^^^may denote a channel at time ^frequency band^, ^ ∈ {1,2, … , ^} , between ^^ and ^^ which may form a matrix of size ^ × ^ withcomplex entries, i.e., ^^ ^×^^ ^^^ ∈ ℂ .

[0034] At time ^ and frequency band ^, it may be assumed that the NE 102-a wantsto transmit message ^^ ^^ ^^^ to user ^^ where = {1,2, ⋯ , ^} while it uses "^ ^^^ ∈ℂ^×^as a The received signal at ^^, #^^^^^, may be written as:#^^ ^^^ = ^^^ ^^^"^^ ^^^^^^ ^^^ + %^^ ^^^, where %^^ ^^^ represents a noise vector at areceiver.

[0035] To improve an achievable rate of the link, in one example, the NE 102-a selects "^^^^^ that maximizes a received signal-to-interference and noise ratio (SINR). Several different configurations may be used for selection of "^^^^^ where some of the configurations may have some knowledge about ^^^^^^.

[0036] The NE 102-a may get knowledge of ^^^^^^ by direct measurement (e.g., in a time duplex division (TDD) mode and assuming reciprocity of a channel), or indirectly using information that one of the UEs 104-a, 104-b, and 104-c sends to the NE 102-a (e.g., in a frequency division duplex (FDD) mode). In the indirect method, a large amount of feedback may be used to send accurate information about ^^^^^^. This may require a large amount of data if there are a large number of antennas or / and large frequency bands.

[0037] In one example, a single time slot is analyzed, but the example may be extended to the examples with more than a single time slot. Without loss of generality, ^^^^^^ may be denoted using ^^^.

[0038] ^^^^^ may be defined as a matrix of size ^ × ^ × ^ which is formed bystacking ^^^ for all frequency bands (e.g., the entries at ^^&', (, ^)^^^ is equal to^^^ &', ()^^^^. In total, each UE 104-a, 104-b, and 104-c may provide feedbackinformation about most recent ^ × ^ × ^ complex numbers to the NE 102-a.

[0039] Various configurations may be used to reduce a rate of required feedback. Two-sided models may include two parts where a first part is deployed at a UE side (e.g., UE 104) and the second part is deployed at a gNB side (e.g., NE 102). The UE and gNB sides may include one or a few neural network (NN) blocks which are trained using data driven approaches. The UE side may be responsible for computing a latentrepresentation of input data (e.g., what needs to be transferred to the gNB) with as low of a number of bits as possible. Receiving what has been transmitted by the UE side, the gNB side reconstructs the information intended to be transmitted to the gNB. An encoding part of the two-sided model (at Node A, e.g., UE) may compute a quantized latent representation of the input data, and a decoding part of the two-sided model (at Node B, e.g., the gNB) may get this latent representation and use it to reconstruct the desired output.

[0040] Figure 3 illustrates an example of a block diagram of a two-sided model 300 (e.g., NN-based model) in accordance with aspects of the present disclosure. The two- sided model 300 includes a node A 302 (e.g., encoder, encoding model, Me) and a node B 304 (e.g., decoder, decoding model, Md). Figure 3 illustrates only one example of a two-sided model 300, while in other examples the location of the encoder and decoder may be swapped. Input data 306 is provided to the node A 302, then there is a latent representation 308 in communications from the node A 302 to the node B 304, and the node B 304 outputs data 310. In some configurations, the node A 302 is located at the UE 104, and the node B 304 is located at the NE 102, but in other configurations the node A 302 and the node B 304 may be located in other devices.

[0041] The exact structure of a UE and a gNB side may vary depending on a particular scheme, but in some embodiment training of an encoder and decoder are not online, meaning that the encoder and the decoder training is based on some criteria and an NN-block of the encoder and the decoder are fixed during an inference phase.

[0042] There may be several methods to train NN modules at UE and gNB sides, including, centralized training, simultaneous training, and separate training. Similarly, updating the two-sided model 300 may be carried out centrally on one entity, on different entities but simultaneously, or separately.

[0043] In a separate training and / or model update, NN modules of the node A 302 (e.g., UE) and the node B 304 (e.g., gNB) parts are trained in different training sessions (e.g., no forward or backpropagation path between the two parts). In separate training, the node A 302 does not need to know ^*and the node B 304 does not need to know ^+.

[0044] To have a model with high generalization capability, one configuration trains the model with as many diverse samples as possible. For example, gathering data frommany different UEs and many different cells to train a model that may be usable in different scenarios. Having a model with high generalization capability simplifies the operation of the system as there may be no need to handle different models for different cases (e.g., reduces overhead related to signaling and model transfer).

[0045] In some scenarios, a model trained for a specific task (e.g., task A) performs better for task A compared to a generalized model trained for task A, task B, and task C. So, if overhead related to model management is handled, it may be beneficial to have multiple models each for each task.

[0046] In one example, there may be two two-sided channel state information (CSI) feedback models: a first model, named ℳ^, is trained only with data collected from a particular UE (e.g., UE-1), and a model, named ℳ^, is trained with datacollected from multiple UEs (e.g., UE-2,…, UE-K). If there are enough training samples for training both models, the performance of ℳ^may be higher than ℳ^if the model for UE-1 is used.

[0047] In some examples found herein there are different schemes for supporting UE-specific models. For two-sided models, one way to support different UEs is to collect data from many UEs and then train one single model using all of the data collected. For a separate training method, the node A 302 (e.g., UE-side node) has an encoder part of the model and the node B 304 (e.g., the network side node) has a decoder part of the model. As the model is trained using data from all UEs, it may be delivered to all UEs and all gNBs. Model management in this example is easier as the network and / or UE needs to deal only with a single model. The drawback, however, is that the performance may be lower than a scheme in which one model is trained for each UE.

[0048] In various examples, data is collected from different UEs, but instead of combining all of them, separate models are trained for different UEs. Assuming ^ different UEs, there may be ^ different decoder models and also ^ different encoder models. In one example, data is collected from different UEs, but instead of combining all of them, separate models are trained for different UEs. Assuming ^ different UEs, in this example, there may be ^ different decoder models and also ^ different encoder models. Each UE then may receive an encoder model and the gNB may use a corresponding decoding part for each UE. For this example, model management may becomplicated and may lead to higher signaling overhead. Moreover, this example may not be scalable if new UEs appear in a network and a training step may need to be performed for each new UE. For a single UE, if the UE changes its location, the statistics of input data may change which may necessitate modification of the encoder and / or decoder model.

[0049] Described herein are various embodiments that may be used to train models that may capture UE properties while having low model management complexity.

[0050] In certain examples, there are ^ datasets, -. = / 〈13 32, 42〉, 6 = 1,2, ⋯ , ^78, 9 =1,2, ⋯ , ^, each for a particular task, e.g., CSI dataset collected from the 9:; UE where9 = 1,2, ⋯ , ^, where 13 and 43 represents the 6:; i :;2 2 nput and desired output of the 9dataset. In some cases, 43could32 be equal to 12. Then, we construct a complete dataset called: -= < -.

[0051] Some configurationsmode A specific models. In a training phase, a training dataset - is used to train ^+>and ^*>as a pair of generalizable encoder and decoder that may be used for all tasks (e.g.,. Next, there may be another NN block, namely ^?@+that may be used as a pre-encoder block. A complete two-sided model may be the concatenation of ^ > >?@+ , ^+ , ^* models as shown in Figure4. It should be noted that, in practice, ^ may bea smaller NN block ^+>and ^*>. For example, ^?@+may include a few convolutional or fully connectedmay be defined as a vector constructed by concatenating all weights ^?@+.

[0052] Figure 4 illustrates an example of a block diagram of another two-sided model 400 in accordance with aspects of the present disclosure. The two-sided model 400 includes a node A 402 (e.g., encoder, encoding model, ^+>), a node B 304 (e.g., decoder, decoding model, ^*>), and a pre-encoder node 406 (e.g., ^?@+). Figure 4 illustrates only onethe two-sided model 400, while in other examples the location of the encoder and decoder may be swapped. Input data 408 is provided to the pre-encoder node 406, then there is a latent representation 410 in communications from the node A 402 to the node B 404, and the node B 404 outputs data 412. In someconfigurations, the node A 402 is located at the UE 104, and the node B 404 is located at the NE 102, but in other configurations the node A 402 and the node B 404 may be located in other devices.

[0053] For the resulted two-sided model 400 ^ models are trained, each with one of the tasks -.(e.g., UE-specific datasets) while the encoder and the decoder blocks are kept constant based on the generalized models developed using dataset -, e.g., weights are frozen to ^+>, and ^*>. As a result, there may be ^ pre-encoder blocks and ^?7@+may be used where 9 ⋯ , ^. Moreover, after this step, there may be "7? for a@+task (e.g., UE) 9 where 9 = 1,2, ⋯ , ^. A complete model for a 9:; task (e.g., UE) maybe: ^7 → > >?@+ ^+ → ^*.training dataset CDEFmay be constructed, where CDEF= / 〈{132, 6 = 1,2, ⋯ , ^}, "7?@+ 〉, 9 = 1,2, ⋯ , ^8. Each sample from this dataset may have twocomponents. The first component may include a set of ^ samples from 132 of one of the tasks (e.g., UEs), e.g., UE 9, and the second component may include a pre-encoder blocks weight vector constructed in a previous step for that cell, e.g., "?7@+. Different ^ samples may be randomly selected from the set of 132 collected for the task 9. This may help construction of larger CDEFdataset.

[0055] Having CDEF, neural network blocks called G?@+, are trained, which act as a model to generate different pre-encoder models for different tasks (e.g., UEs). Morespecifically, C 3DEF is used to train G?@+ such that it gets {12, 6 = 1,2, ⋯ , ^} from one task(e.g., UE) as the input and predicts "?7@+for that task (e.g., UE). G?@+may be the same for all tasks (e.g., UEs).

[0056] In one example of training models, the encoder and decoder models are available at separate entities. In such an example there are ^ node As and H node Bs. There are also two nodes, one from node A, called IJ, and one from node Be, called IK, which have access to data from all node As and node Bs, respectively.

[0057] IJmay have access to K training datasets where the 9:;training data set- = / 〈17 , 47〉, ^ = 1,2, ⋯ , ^ 8 repre 77 ^ ^ 7 sent ^7 samples where 1^ is an input of the two-sided model 400 and 47^ is a desired output of the two-sided model collected at the 9:;node A. This training dataset (e.g., CSI dataset) may be collected and / or measured from the environment.

[0058] There may be different methods to train the models. In one method, a decoder is constructed and then an encoder model is constructed (e.g., a node B first configuration). Other configurations may be used for training the encoder module first.

[0059] In certain examples, IJis assumed or individual node As first transmit -7toIK . IK then may combine all datasets and construct - = ⋃7=^,^,⋯,^ -.. IK may thenconsider a NN structure as its local encoder model ^ℳM+^ and may assume an NN structure for a decoder part ^ℳ*). It should be noted that ℳM+is not an actual encoder model to be used in node As. It is one model that IKmay consider for itself to be able to train the second part, e.g., ℳ*.

[0060] IKmay then use the dataset - to train a local two-sided model having ℳM+and ℳ*, such that ℳ*NℳM+^17^^O becomes as close as possible to 47^ for all samples existing in -. ℳM+∗and ℳ*∗may be used to refer to trained models. IKmay then fix ℳ*to ℳ∗ M 7* and then uses samples of -. to train a local two-sided model including ℳ + andthat ℳ*∗NℳM+7^17^^O becomes as close as possible to 47^ for all samples existing +7∗in -. ℳMmay be used to refer to a local encoder model for each dataset.

[0061] IKmay then use the trained local two-sided model to generate a trainingdataset QSS⃖ = / 〈1^ , T^ = ℳM ∗+^1^^〉, ^ = 1,2, ⋯ , ^8 for all or subset of 1^U existing indataset -. The T^may represent a latent representation of input data 1Vbased on a trained local model of encoder,Wℳ+∗.

[0062] IK may also generate K other training datasets QSS7⃖ = X〈17 7^ , T^ =∗ℳM 7^17^^〉 , ^ = 1,2, ⋯ , ^Y for all or a subset of 17^U existing.. I may thensend QSS⃖ and all QSS7⃖ for 9 = 1,2, ⋯ , ^ to the IJ.

[0063] IJmay then: use QSS⃖ to train the encoder model ℳ+such that ℳ+>^1^^ becomes as close as possible to T^for all samples in QSS⃖, use ℳ+>to refer to this trained model, use QSS7⃖ to train pre-encoder model ^?7@+such that ℳ+>^^?7@+^17^^^ becomes as close as possible to T7^ for all samples in QSS7⃖ - this may result in ^ trained pre-encoder?7∗ @+ ?7∗ @+mine "?7∗ models referred to as ^ , serialize the weights of ^ to deter@+,construct dataset CDEF = / 〈{132, 6 = 1,2, ⋯ , ^}, "7?@+ 〉, 9 = 1,2, ⋯ , ^8 where the firstcomponent of each sample is a set ^ samples from 13QSS⃖ , and / orC to train neural network blocks, 3DEF called G?@+, such that G?@+^{12, 6 = 1,2, ⋯ , ^}^here 13∗ w2sampled from QSS7⃖ become as close as possible to "?7@+. This may conclude the training process which may result in ℳ*∗for the decoder model and the two models of ℳ>and G?@+which are used for construction of the encoder model for each UE.should be noted that the process may to improve an accuracy of trained model.

[0064] In some examples, there may be an inference phase. In such examples, ^+>, ^*>, and G?@+may be trained and available, the node B (e.g., the gNB) may receive ^*>decoder model which may be used as a second part of the two-sided model,all node As (e.g., the UEs) may receive ^+>as the common part of the encoder model and also CDEFas a generator model for pre-encoder blocks. In some implementations, node As (e.g., the UEs) may only receive ^+>as a common part of the encoder model.

[0065] In one example, each node A, e.g., the 9:;node A, may then use CDEFto determine a good "?7@+based on its own 13Z. Using "?7@+the corresponding node A maythen construct a ^7 model. The complete encoder m 7 >?@+ odel may then be ^?@+ → ^*.

[0066] Node A may be able to regularly update ^?7@+based on an input data it receives, e.g., 13Z, to have a better match model for current input statistics. If node A updates its model, it may inform the other side that the encoder model has been modified. It may be useful for model management and also if the other side wants to make changes in the decoding procedure.

[0067] In one implementation, the 9:;node A transmits a few samples from its own input data 13Z to another node (e.g., another node on the NW, on the UE vendor side, on the operator side) which has access to CDEFas a generator model for a pre-encoder block. That node may then use CDEFto determine a good "?7@+and may transmit it back to node A. Node A may then use the received data to constructs the ^?7@+model. Thecomplete encoder model may then be ^7 > 7?@+ → ^*. Node A may update ^?@+ bysending some samples to that node and "?7@+using which it may update ^?7@+.

[0068] It should be noted that any embodiments described herein are not limited to their particular application. The embodiments may be used in any other task specific use case that has access to different models developed for different task and may provide an advantage over a single model developed to support all tasks. It should also be noted that the training phase may have an assumption a node B does not know node A models and models may be delivered using a dataset transfer.

[0069] Figure 5 illustrates an example of a UE 500 in accordance with aspects of the present disclosure. The UE 500 may include a processor 502, a memory 504, a controller 506, and a transceiver 508. The processor 502, the memory 504, the controller 506, or the transceiver 508, or various combinations thereof or various components thereof may be examples of means for performing various aspects of the present disclosure as described herein. These components may be coupled (e.g., operatively, communicatively, functionally, electronically, electrically) via one or more interfaces.

[0070] The processor 502, the memory 504, the controller 506, or the transceiver 508, or various combinations or components thereof may be implemented in hardware (e.g., circuitry). The hardware may include a processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), or other programmable logic device, or any combination thereof configured as or otherwise supporting a means for performing the functions described in the present disclosure.

[0071] The processor 502 may include an intelligent hardware device (e.g., a general-purpose processor, a DSP, a CPU, an ASIC, a field programmable gate array (FPGA), or any combination thereof). In some implementations, the processor 502 may be configured to operate the memory 504. In some other implementations, the memory 504 may be integrated into the processor 502. The processor 502 may be configured to execute computer-readable instructions stored in the memory 504 to cause the UE 500 to perform various functions of the present disclosure.

[0072] The memory 504 may include volatile or non-volatile memory. The memory 504 may store computer-readable, computer-executable code including instructions when executed by the processor 502 cause the UE 500 to perform various functions described herein. The code may be stored in a non-transitory computer-readable medium such the memory 504 or another type of memory. Computer-readable media includes both non-transitory computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A non-transitory storage medium may be any available medium that may be accessed by a general-purpose or special-purpose computer.

[0073] In some implementations, the processor 502 and the memory 504 coupled with the processor 502 may be configured to cause the UE 500 to perform one or more of the functions described herein (e.g., executing, by the processor 502, instructions stored in the memory 504). For example, the processor 502 may support wireless communication at the UE 500 in accordance with examples as disclosed herein. For example, the processor 502 coupled with the memory 504 may be configured to cause the UE 500 to determine a first set of one or more parameters for an encoder model of a two-sided model. The UE 500 may also determine a first set of information comprising samples that represent an input to the two-sided model. The UE 500 may update the encoder model according to one or more of a third set of one or more parameters associated with a pre-encoder model, a first message from another apparatus, or a second set of one or more parameters for a pre-encoder generator model, wherein the third set of one or more parameters is based at least in part on the first set of information. The UE 500 may also encode data according to the updated encoder model and the first set of information. The UE 500 may transmit the encoded data to a second apparatus.

[0074] The controller 506 may manage input and output signals for the UE 500. The controller 506 may also manage peripherals not integrated into the UE 500. In some implementations, the controller 506 may utilize an operating system such as iOS®, ANDROID®, WINDOWS®, or other operating systems. In some implementations, the controller 506 may be implemented as part of the processor 502.

[0075] In some implementations, the UE 500 may include at least one transceiver 508. In some other implementations, the UE 500 may have more than one transceiver508. The transceiver 508 may represent a wireless transceiver. The transceiver 508 may include one or more receiver chains 510, one or more transmitter chains 512, or a combination thereof.

[0076] A receiver chain 510 may be configured to receive signals (e.g., control information, data, packets) over a wireless medium. For example, the receiver chain 510 may include one or more antennas for receive the signal over the air or wireless medium. The receiver chain 510 may include at least one amplifier (e.g., a low-noise amplifier (LNA)) configured to amplify the received signal. The receiver chain 510 may include at least one demodulator configured to demodulate the receive signal and obtain the transmitted data by reversing the modulation technique applied during transmission of the signal. The receiver chain 510 may include at least one decoder for decoding the processing the demodulated signal to receive the transmitted data.

[0077] A transmitter chain 512 may be configured to generate and transmit signals (e.g., control information, data, packets). The transmitter chain 512 may include at least one modulator for modulating data onto a carrier signal, preparing the signal for transmission over a wireless medium. The at least one modulator may be configured to support one or more techniques such as amplitude modulation (AM), frequency modulation (FM), or digital modulation schemes like phase-shift keying (PSK) or quadrature amplitude modulation (QAM). The transmitter chain 512 may also include at least one power amplifier configured to amplify the modulated signal to an appropriate power level suitable for transmission over the wireless medium. The transmitter chain 512 may also include one or more antennas for transmitting the amplified signal into the air or wireless medium.

[0078] Figure 6 illustrates an example of a processor 600 in accordance with aspects of the present disclosure. The processor 600 may be an example of a processor configured to perform various operations in accordance with examples as described herein. The processor 600 may include a controller 602 configured to perform various operations in accordance with examples as described herein. The processor 600 may optionally include at least one memory 604, which may be, for example, an L1 / L2 / L3 cache. Additionally, or alternatively, the processor 600 may optionally include one or more arithmetic-logic units (ALUs) 606. One or more of these components may be inelectronic communication or otherwise coupled (e.g., operatively, communicatively, functionally, electronically, electrically) via one or more interfaces (e.g., buses).

[0079] The processor 600 may be a processor chipset and include a protocol stack (e.g., a software stack) executed by the processor chipset to perform various operations (e.g., receiving, obtaining, retrieving, transmitting, outputting, forwarding, storing, determining, identifying, accessing, writing, reading) in accordance with examples as described herein. The processor chipset may include one or more cores, one or more caches (e.g., memory local to or included in the processor chipset (e.g., the processor 600) or other memory (e.g., random access memory (RAM), read-only memory (ROM), dynamic RAM (DRAM), synchronous dynamic RAM (SDRAM), static RAM (SRAM), ferroelectric RAM (FeRAM), magnetic RAM (MRAM), resistive RAM (RRAM), flash memory, phase change memory (PCM), and others).

[0080] The controller 602 may be configured to manage and coordinate various operations (e.g., signaling, receiving, obtaining, retrieving, transmitting, outputting, forwarding, storing, determining, identifying, accessing, writing, reading) of the processor 600 to cause the processor 600 to support various operations in accordance with examples as described herein. For example, the controller 602 may operate as a control unit of the processor 600, generating control signals that manage the operation of various components of the processor 600. These control signals include enabling or disabling functional units, selecting data paths, initiating memory access, and coordinating timing of operations.

[0081] The controller 602 may be configured to fetch (e.g., obtain, retrieve, receive) instructions from the memory 604 and determine subsequent instruction(s) to be executed to cause the processor 600 to support various operations in accordance with examples as described herein. The controller 602 may be configured to track memory address of instructions associated with the memory 604. The controller 602 may be configured to decode instructions to determine the operation to be performed and the operands involved. For example, the controller 602 may be configured to interpret the instruction and determine control signals to be output to other components of the processor 600 to cause the processor 600 to support various operations in accordance with examples as described herein. Additionally, or alternatively, the controller 602 may be configured to manage flow of data within the processor 600. The controller 602 maybe configured to control transfer of data between registers, arithmetic logic units (ALUs), and other functional units of the processor 600.

[0082] The memory 604 may include one or more caches (e.g., memory local to or included in the processor 600 or other memory, such RAM, ROM, DRAM, SDRAM, SRAM, MRAM, flash memory, etc. In some implementations, the memory 604 may reside within or on a processor chipset (e.g., local to the processor 600). In some other implementations, the memory 604 may reside external to the processor chipset (e.g., remote to the processor 600).

[0083] The memory 604 may store computer-readable, computer-executable code including instructions that, when executed by the processor 600, cause the processor 600 to perform various functions described herein. The code may be stored in a non- transitory computer-readable medium such as system memory or another type of memory. The controller 602 and / or the processor 600 may be configured to execute computer-readable instructions stored in the memory 604 to cause the processor 600 to perform various functions. For example, the processor 600 and / or the controller 602 may be coupled with or to the memory 604, the processor 600, the controller 602, and the memory 604 may be configured to perform various functions described herein. In some examples, the processor 600 may include multiple processors and the memory 604 may include multiple memories. One or more of the multiple processors may be coupled with one or more of the multiple memories, which may, individually or collectively, be configured to perform various functions herein.

[0084] The one or more ALUs 606 may be configured to support various operations in accordance with examples as described herein. In some implementations, the one or more ALUs 606 may reside within or on a processor chipset (e.g., the processor 600). In some other implementations, the one or more ALUs 606 may reside external to the processor chipset (e.g., the processor 600). One or more ALUs 606 may perform one or more computations such as addition, subtraction, multiplication, and division on data. For example, one or more ALUs 606 may receive input operands and an operation code, which determines an operation to be executed. One or more ALUs 606 be configured with a variety of logical and arithmetic circuits, including adders, subtractors, shifters, and logic gates, to process and manipulate the data according to the operation. Additionally, or alternatively, the one or more ALUs 606 may support logicaloperations such as AND, OR, exclusive-OR (XOR), not-OR (NOR), and not-AND (NAND), enabling the one or more ALUs 606 to handle conditional operations, comparisons, and bitwise operations.

[0085] The processor 600 may support wireless communication in accordance with examples as disclosed herein. The processor 600 may be configured to or operable to support a means for: determining a first set of one or more parameters for an encoder model of a two-sided model, determining a first set of information comprising samples that represent an input to the two-sided model, updating the encoder model according to one or more of a third set of one or more parameters associated with a pre-encoder model, a first message from another apparatus, or a second set of one or more parameters for a pre-encoder generator model, wherein the third set of one or more parameters is based at least in part on the first set of information, encoding data according to the updated encoder model and the first set of information, and transmitting the encoded data to a second apparatus.

[0086] Figure 7 illustrates an example of a NE 700 in accordance with aspects of the present disclosure. The NE 700 may include a processor 702, a memory 704, a controller 706, and a transceiver 708. The processor 702, the memory 704, the controller 706, or the transceiver 708, or various combinations thereof or various components thereof may be examples of means for performing various aspects of the present disclosure as described herein. These components may be coupled (e.g., operatively, communicatively, functionally, electronically, electrically) via one or more interfaces.

[0087] The processor 702, the memory 704, the controller 706, or the transceiver 708, or various combinations or components thereof may be implemented in hardware (e.g., circuitry). The hardware may include a processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), or other programmable logic device, or any combination thereof configured as or otherwise supporting a means for performing the functions described in the present disclosure.

[0088] The processor 702 may include an intelligent hardware device (e.g., a general-purpose processor, a DSP, a CPU, an ASIC, an FPGA, or any combination thereof). In some implementations, the processor 702 may be configured to operate the memory 704. In some other implementations, the memory 704 may be integrated intothe processor 702. The processor 702 may be configured to execute computer-readable instructions stored in the memory 704 to cause the NE 700 to perform various functions of the present disclosure. For example, the processor 702 coupled with the memory 704 may be configured to cause the NE 700 to: determine a set of parameters for a decoder model of a two-sided model, receive an encoded data from a first apparatus, and determine a model output of the two-sided model based on the encoded data and the decoder model.

[0089] The memory 704 may include volatile or non-volatile memory. The memory 704 may store computer-readable, computer-executable code including instructions when executed by the processor 702 cause the NE 700 to perform various functions described herein. The code may be stored in a non-transitory computer-readable medium such the memory 704 or another type of memory. Computer-readable media includes both non-transitory computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A non-transitory storage medium may be any available medium that may be accessed by a general-purpose or special-purpose computer.

[0090] In some implementations, the processor 702 and the memory 704 coupled with the processor 702 may be configured to cause the NE 700 to perform one or more of the functions described herein (e.g., executing, by the processor 702, instructions stored in the memory 704). For example, the processor 702 may support wireless communication at the NE 700 in accordance with examples as disclosed herein.

[0091] The controller 706 may manage input and output signals for the NE 700. The controller 706 may also manage peripherals not integrated into the NE 700. In some implementations, the controller 706 may utilize an operating system such as iOS®, ANDROID®, WINDOWS®, or other operating systems. In some implementations, the controller 706 may be implemented as part of the processor 702.

[0092] In some implementations, the NE 700 may include at least one transceiver 708. In some other implementations, the NE 700 may have more than one transceiver 708. The transceiver 708 may represent a wireless transceiver. The transceiver 708 may include one or more receiver chains 710, one or more transmitter chains 712, or a combination thereof.

[0093] A receiver chain 710 may be configured to receive signals (e.g., control information, data, packets) over a wireless medium. For example, the receiver chain 710 may include one or more antennas for receive the signal over the air or wireless medium. The receiver chain 710 may include at least one amplifier (e.g., a low-noise amplifier (LNA)) configured to amplify the received signal. The receiver chain 710 may include at least one demodulator configured to demodulate the receive signal and obtain the transmitted data by reversing the modulation technique applied during transmission of the signal. The receiver chain 710 may include at least one decoder for decoding the processing the demodulated signal to receive the transmitted data.

[0094] A transmitter chain 712 may be configured to generate and transmit signals (e.g., control information, data, packets). The transmitter chain 712 may include at least one modulator for modulating data onto a carrier signal, preparing the signal for transmission over a wireless medium. The at least one modulator may be configured to support one or more techniques such as amplitude modulation (AM), frequency modulation (FM), or digital modulation schemes like phase-shift keying (PSK) or quadrature amplitude modulation (QAM). The transmitter chain 712 may also include at least one power amplifier configured to amplify the modulated signal to an appropriate power level suitable for transmission over the wireless medium. The transmitter chain 712 may also include one or more antennas for transmitting the amplified signal into the air or wireless medium.

[0095] Figure 8 illustrates a flowchart of a method 800 in accordance with aspects of the present disclosure. The operations of the method 800 may be implemented by a first apparatus (e.g., UE) as described herein. In some implementations, a UE 500 may execute a set of instructions to control the function elements of a processor to perform the described functions.

[0096] At 802, the method may include determining a first set of one or more parameters for an encoder model of a two-sided model. The operations of 802 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 802 may be performed by a UE as described with reference to Figure 5.

[0097] At 804, the method may include determining a first set of information comprising samples that represent an input to the two-sided model. The operations of804 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 804 may be performed by a UE as described with reference to Figure 5.

[0098] At 806, the method may include updating the encoder model according to one or more of a third set of one or more parameters associated with a pre-encoder model, a first message from another apparatus, or a second set of one or more parameters for a pre-encoder generator model, wherein the third set of one or more parameters is based at least in part on the first set of information. The operations of 806 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 806 may be performed by a UE as described with reference to Figure 5.

[0099] At 808, the method may include encoding data according to the updated encoder model and the first set of information. The operations of 808 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 808 may be performed by a UE as described with reference to Figure 5.

[0100] At 810, the method may include transmitting the encoded data to a second apparatus. The operations of 810 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 810 may be performed by a UE as described with reference to Figure 5.

[0101] Figure 9 illustrates a flowchart of another method 900 in accordance with aspects of the present disclosure. The operations of the method 900 may be implemented by a second apparatus (e.g., NE) as described herein. In some implementations, a NE 700 may execute a set of instructions to control the function elements of a processor to perform the described functions.

[0102] At 902, the method may include determining a set of parameters for a decoder model of a two-sided model. The operations of 902 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 902 may be performed by a NE as described with reference to Figure 7.

[0103] At 904, the method may include receiving an encoded data from a first apparatus. The operations of 904 may be performed in accordance with examples asdescribed herein. In some implementations, aspects of the operations of 904 may be performed by a NE as described with reference to Figure 7.

[0104] At 906, the method may include determining a model output of the two- sided model based on the encoded data and the decoder model. The operations of 906 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 906 may be performed by a NE as described with reference to Figure 7.

[0105] Figure 10 illustrates a flowchart of another method 1000 in accordance with aspects of the present disclosure. The operations of the method 1000 may be implemented by a second apparatus (e.g., NE) as described herein. In some implementations, a NE 700 may execute a set of instructions to control the function elements of a processor to perform the described functions.

[0106] At 1002, the method may include determining a set of training datasets, wherein each training dataset of the set of training datasets comprising samples representing an input to a two-sided model, an output of the two-sided model, or a combination thereof. The operations of 1002 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 1002 may be performed by a NE as described with reference to Figure 7.

[0107] At 1004, the method may include determining a first set of parameters and a second set of parameters using the set of training datasets, wherein the first set of parameters is associated with a first model and the second set of parameters is associated with a second model. The operations of 1004 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 1004 may be performed by a NE as described with reference to Figure 7.

[0108] At 1006, the method may include determining a third set of parameters for each training dataset of the set of training datasets based on the second model, wherein the third set of parameters comprises information for a third model associated with each training dataset of the set of training datasets. The operations of 1006 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 1006 may be performed by a NE as described with reference to Figure 7.

[0109] At 1008, the method may include determining a first feedback dataset based on the set of training datasets and the first set of parameters. The operations of 1008 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 1008 may be performed by a NE as described with reference to Figure 7.

[0110] At 1010, the method may include determining a set of feedback datasets, wherein samples of each feedback datasets of the set of feedback datasets is based on a corresponding training dataset of the set of training datasets, and associated parameters of the third set of parameters, or a combination thereof. The operations of 1010 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 1010 may be performed by a NE as described with reference to Figure 7.

[0111] At 1012, the method may include transmitting the first feedback dataset and the set of feedback datasets to a first apparatus. The operations of 1012 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 1012 may be performed by a NE as described with reference to Figure 7.

[0112] Figure 11 illustrates a flowchart of another method 1100 in accordance with aspects of the present disclosure. The operations of the method 1100 may be implemented by a first apparatus (e.g., UE) as described herein. In some implementations, a UE 500 may execute a set of instructions to control the function elements of a processor to perform the described functions.

[0113] At 1102, the method may include determining a first dataset, a set of datasets, or a combination thereof, wherein samples of the first dataset, each dataset of the set of datasets, or a combination thereof comprises samples representing an input, a desired output of a model, or a combination thereof. The operations of 1102 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 1102 may be performed by a UE as described with reference to Figure 5.

[0114] At 1104, the method may include determining a first set of parameters based on the first dataset, wherein the first set of parameters comprises information for a first model, and determining a second set of parameters for each dataset of the set ofdatasets, wherein each second set of parameters comprises information for a pre- encoder model; or a combination thereof. The operations of 1104 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 1104 may be performed by a UE as described with reference to Figure 5.

[0115] At 1106, the method may include determining a third dataset, wherein samples of the third dataset represent a weight vector comprising weights for a corresponding pre-encoder model and a set of samples from input data associated with a dataset of the set of datasets associated with the corresponding pre-encoder model. The operations of 1106 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 1106 may be performed by a UE as described with reference to Figure 5.

[0116] At 1108, the method may include determining a third set of parameters based on a third dataset, wherein the third set of parameters comprises information associated with a second model. The operations of 1108 may be performed in accordance with examples as described herein. In some implementations, aspects of the operations of 1108 may be performed by a UE as described with reference to Figure 5.

[0117] It should be noted that the methods described herein describe possible implementations, and that the operations and the steps may be rearranged or otherwise modified and that other implementations are possible.

[0118] The description herein is provided to enable a person having ordinary skill in the art to make or use the disclosure. Various modifications to the disclosure will be apparent to a person having ordinary skill in the art, and the generic principles defined herein may be applied to other variations without departing from the scope of the disclosure. Thus, the disclosure is not limited to the examples and designs described herein but is to be accorded the broadest scope consistent with the principles and novel features disclosed herein.

Claims

CLAIMS What is claimed is:

1. A first apparatus, comprising: at least one memory; and at least one processor coupled with the at least one memory and configured to cause the first apparatus to: determine a first set of one or more parameters for an encoder model of a two-sided model; determine a first set of information comprising samples that represent an input to the two-sided model; update the encoder model according to one or more of a third set of one or more parameters associated with a pre-encoder model, a first message from another apparatus, or a second set of one or more parameters for a pre-encoder generator model, wherein the third set of one or more parameters is based at least in part on the first set of information; encode data according to the updated encoder model and the first set of information; and transmit the encoded data to a second apparatus.

2. The first apparatus of claim 1, wherein the at least one processor is configured to cause the first apparatus to determine the encoder model based on the first set of one or more parameters and the third set of one or more parameters.

3. The first apparatus of claim 1, wherein the at least one processor is configured to cause the first apparatus to transmit a second message to another apparatus, and the second message comprises a subset of the samples from the first set of information.

4. The first apparatus of claim 1, wherein the encoder model is updated based on a periodic event, receiving a triggering message from another apparatus, observing a shift in statistics of the first set of information, observing a shift in statistics ofa second set of information, or a combination thereof, and wherein the second set of information is based on parameters of an environment or parameters of the first apparatus.

5. The first apparatus of claim 1, wherein, in response to the encoder model being updated, the at least one processor is configured to cause the first apparatus to transmit a triggering message to another apparatus indicating that the pre- encoder model is modified.

6. The first apparatus of claim 1, wherein the first set of information is based on a channel data representation.

7. The first apparatus of claim 6, wherein the channel data representation is based on reception of at least one reference signal from the second apparatus.

8. The first apparatus of claim 6, wherein the channel data representation is based on different transmit (TX)-receive (RX) pairs over different frequency bands or different time slots or TX-RX pair transformations in other domains.

9. The first apparatus of claim 1, wherein the first set of one or more parameters, the second set of one or more parameters, or a combination thereof is received from another apparatus.

10. The first apparatus of claim 1, wherein the first apparatus comprises a user equipment (UE).

11. The first apparatus of claim 1, wherein the second apparatus comprises a base station.

12. A second apparatus, comprising: at least one memory; andat least one processor coupled with the at least one memory and configured to cause the second apparatus to: determine a set of parameters for a decoder model of a two-sided model; receive an encoded data from a first apparatus; and determine a model output of the two-sided model based on the encoded data and the decoder model.

13. The second apparatus of claim 12, wherein the set of parameters is received from another apparatus.

14. The second apparatus of claim 12, wherein the second apparatus comprises a base station.

15. A second apparatus, comprising: at least one memory; and at least one processor coupled with the at least one memory and configured to cause the second apparatus to: determine a set of training datasets, wherein each training dataset of the set of training datasets comprising samples representing an input to a two-sided model, an output of the two-sided model, or a combination thereof; determine a first set of parameters and a second set of parameters using the set of training datasets, wherein the first set of parameters is associated with a first model and the second set of parameters is associated with a second model; determine a third set of parameters for each training dataset of the set of training datasets based on the second model, wherein the third set of parameters comprises information for a third model associated with each training dataset of the set of training datasets; determine a first feedback dataset based on the set of training datasets and the first set of parameters;determine a set of feedback datasets, wherein samples of each feedback datasets of the set of feedback datasets is based on a corresponding training dataset of the set of training datasets, and associated parameters of the third set of parameters, or a combination thereof; and transmit the first feedback dataset and the set of feedback datasets to a first apparatus.

16. The second apparatus of claim 15, wherein the input is based on a channel data representation.

17. The second apparatus of claim 16, wherein at least part of the set of training datasets is received from another apparatus.

18. A first apparatus, comprising: at least one memory; and at least one processor coupled with the at least one memory and configured to cause the first apparatus to: determine a first dataset, a set of datasets, or a combination thereof, wherein samples of the first dataset, each dataset of the set of datasets, or a combination thereof comprises samples representing an input, a desired output of a model, or a combination thereof; determine a first set of parameters based on the first dataset, wherein the first set of parameters comprises information for a first model, and determine a second set of parameters for each dataset of the set of datasets, wherein each second set of parameters comprises information for a pre-encoder model; or a combination thereof; determine a third dataset, wherein samples of the third dataset represent a weight vector comprising weights for a corresponding pre-encoder model and a set of samples from input data associated with a dataset of the set ofdatasets associated with the corresponding pre-encoder model; and determine a third set of parameters based on a third dataset, wherein the third set of parameters comprises information associated with a second model.

19. The first apparatus of claim 18, wherein the input is based on a channel data representation.

20. The first apparatus of claim 18, wherein at least part of the first dataset and the set of datasets is received from another apparatus.

Images