Size-based neural network selection for autoencoder-based communication

Abstract

Methods, systems, and apparatuses for wireless communication are described. In some wireless communication systems, the apparatus may implement multiple autoencoders for communication. The wireless apparatus may select the autoencoder for communication based on a size parameter for the message. For example, a user equipment (UE) may receive a permission from a base station indicating the size parameter for a communication message. The UE and the base station may determine the NN-based encoder corresponding to the size parameter from a set of neural network (NN)-based encoders configured at the UE. The UE may then communicate the message with the base station based on the permission and the determined NN encoder. In some examples, the UE and the base station may determine a number of resource segments from a set of resources allocated for communication, and may determine the corresponding NN-based encoder for different resource segments.

Description

Technical Field

[0001] The following relates to wireless communication, including size-based neural network (NN) selection for autoencoder-based communication. Background Technology

[0002] Wireless communication systems are widely deployed to provide various types of communication content such as voice, video, packet data, messaging, and broadcasting. These systems may be able to support communication with multiple users by sharing available system resources (e.g., time, frequency, and power). Examples of such multiple access systems include fourth-generation (4G) systems such as Long Term Evolution (LTE), LTE-A Advanced (LTE-A), or LTE-A Pro systems, and fifth-generation (5G) systems, which may be referred to as New Radio (NR). These systems may employ technologies such as Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), Orthogonal Frequency Division Multiple Access (OFDMA), or Discrete Fourier Transform Extended Orthogonal Frequency Division Multiple Access (DFT-S-OFDM). A wireless multiple access communication system may include one or more base stations or one or more network access nodes, each of which simultaneously supports communication with multiple communication devices, which may also be referred to as User Equipment (UE).

[0003] In some wireless communication systems, wireless devices (e.g., UEs or base stations) may utilize autoencoders to encode, decode, or both of transmissions. However, autoencoders can be optimized or otherwise trained based on the specific message size used for communication and may perform relatively poorly for other message sizes. Therefore, wireless devices implementing autoencoders may experience inefficient modulation techniques, large signaling overhead, or both, depending on the size of the communicated message. Furthermore, in cases where messages are transmitted in a non-coherent manner, channel estimation may not be usable to aid message demodulation, leading to potentially unreliable demodulation when using autoencoders. Summary of the Invention

[0004] The described technology relates to improved methods, systems, devices, and apparatuses for supporting size-based neural network (NN) selection for autoencoder-based communication. Typically, the described technology provides a user equipment (UE) and a base station with the ability to select a specific NN for communicating a message from a set of multiple configured NNs based on a size parameter for the message. In some cases, the message may be communicated in an incoherent transmission (e.g., a transmission lacking a reference signal, such as a demodulation reference signal (DMRS) for channel estimation). The UE may be configured or pre-configured with a set of NN-based encoders. The UE may receive a permission from the base station indicating a size parameter (e.g., transport block (TB) size or another size parameter) for communicating the message. Based on the size parameter, the UE and the base station may each determine the NN-based encoder (i.e., the same NN-based encoder from the configured set of NN-based encoders). In some cases, multiple neural network-based encoders may be associated with a single size parameter. In some such cases, the base station may explicitly indicate a NN-based encoder using an NN-based encoder index included in the permission. In some other such cases, the UE and the base station may implicitly determine a NN-based encoder based on one or more transmission parameters.

[0005] The UE and base station can communicate messages based on a license and a determined NN-based encoder. For example, one wireless device can encode a message using an NN-based encoder for transmission, and another wireless device can determine an NN-based decoder for decoding the message. In some cases, the NN-based decoder may correspond to the determined NN-based encoder. In other cases, the NN-based decoder may be determined based on an NN-based decoder index indicated by a license. In some examples, separate NN-based encoders and / or decoders can be determined for different segments of allocated resources. For example, a wireless device (e.g., a UE or base station) may determine different segments of allocated resources based on corresponding size parameters of the segments.

[0006] A method for wireless communication at a UE is described. The method may include: receiving from a base station a permission indicating size parameters for communication messages; determining an NN-based encoder corresponding to the size parameters from a set of NN-based encoders configured at the UE; and communicating messages with the base station based on the permission and the determined NN-based encoder.

[0007] An apparatus for wireless communication at a UE is described. The apparatus may include a processor, a memory coupled to the processor, and instructions stored in the memory. The instructions can be executed by the processor to cause the apparatus to: receive a permission from a base station indicating size parameters for a communication message; determine an NN-based encoder corresponding to the size parameters from a set of NN-based encoders configured at the UE; and communicate the message with the base station according to the permission and based on the determined NN-based encoder.

[0008] Another apparatus for wireless communication at a UE is described. The apparatus may include components for performing the following operations: receiving a permission from a base station indicating size parameters for communication messages; determining an NN-based encoder corresponding to the size parameters from a set of NN-based encoders configured at the UE; and communicating messages with the base station based on the permission and the determined NN-based encoder.

[0009] A non-transitory computer-readable medium is described, storing code for wireless communication at a UE. The code may include instructions executable by a processor to: receive a permission from a base station indicating size parameters for communication messages; determine an NN-based encoder corresponding to the size parameters from a set of NN-based encoders configured at the UE; and communicate messages with the base station based on the permission and the determined NN-based encoder.

[0010] In some examples of the methods, apparatuses and nontransitory computer-readable media described herein, communication may include operations, features, components or instructions for communicating messages in incoherent transmissions, which include messages and do not have a reference signal associated with the messages for channel estimation.

[0011] In some examples of the methods, apparatuses and non-transitory computer-readable media described herein, determination may include determining a subset of a set of NN-based encoders corresponding to size parameters and determining the operation, features, components or instructions of an NN-based encoder from the subset of the NN-based encoder set based on an NN-based encoder index indicated by a license.

[0012] In some examples of the methods, apparatuses and non-transitory computer-readable media described herein, determination may include determining a subset of a set of NN-based encoders corresponding to size parameters and determining the operation, features, components or instructions of the NN-based encoders from the subset of the NN-based encoder set based on transmission parameters.

[0013] In some examples of the methods, apparatuses, and non-transitory computer-readable media described herein, transmission parameters include bandwidth indicators, modulation schemes, channel coding rates, or combinations thereof.

[0014] Some examples of the methods, apparatuses, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for receiving configuration messages from a base station that indicate a set of NN-based encoders and one or more size parameters corresponding to each NN-based encoder in the set of NN-based encoders.

[0015] In some examples of the methods, apparatuses and non-transitory computer-readable media described herein, the UE may be pre-configured with a set of NN-based encoders and one or more size parameters corresponding to each NN-based encoder in the set of NN-based encoders.

[0016] In some examples of the methods, apparatuses and non-transitory computer-readable media described herein, the size parameters correspond to the transport block (TB) size for the message, the number of resource elements (REs) for transmitting the message, the number of physical resource blocks (PRBs) for transmitting the message, the number of orthogonal frequency domain multiplexing (OFDM) symbols for transmitting the message, the number of coded bits for transmitting the message, or combinations thereof.

[0017] In some examples of the methods, apparatuses, and non-transitory computer-readable media described herein, communication may include operations, features, components, or instructions for modulating a message using a determined NN-based encoder and transmitting the modulated message to a base station.

[0018] Some examples of the methods, apparatuses, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for determining a NN-based decoder corresponding to the determined NN-based encoder. In some examples of the methods, apparatuses, and non-transitory computer-readable media described herein, communication may include operations, features, components, or instructions for receiving messages from a base station and demodulating the messages using the determined NN-based decoder.

[0019] In some examples of the methods, apparatuses and non-transitory computer-readable media described herein, the NN-based decoder may be determined based on an NN-based decoder index indicated by a license.

[0020] In some examples of the methods, apparatuses and non-transitory computer-readable media described herein, licenses include dynamic uplink licenses, configured licenses, dynamic downlink licenses, semi-persistent scheduling (SPS) configurations or combinations thereof.

[0021] A method for wireless communication at a base station is described. The method may include: sending a permission to a UE indicating size parameters for communication messages; determining an NN-based encoder corresponding to the size parameters from a set of NN-based encoders configured for the UE; and communicating messages with the UE based on the permission and the determined NN-based encoder.

[0022] An apparatus for wireless communication at a base station is described. The apparatus may include a processor, a memory coupled to the processor, and instructions stored in the memory. The instructions can be executed by the processor to cause the apparatus to: send a permission to a UE indicating size parameters for communication messages; determine an NN-based encoder corresponding to the size parameters from a set of NN-based encoders configured for the UE; and communicate messages with the UE according to the permission and based on the determined NN-based encoder.

[0023] Another apparatus for wireless communication at a base station is described. The apparatus may include components for performing the following operations: sending a permission to a UE indicating size parameters for communication messages; determining an NN-based encoder corresponding to the size parameters from a set of NN-based encoders configured for the UE; and communicating messages with the UE based on the permission and the determined NN-based encoder.

[0024] A non-transitory computer-readable medium is described, storing code for wireless communication at a base station. The code may include instructions executable by a processor to: send a permission to a UE indicating size parameters for communication messages; determine an NN-based encoder corresponding to the size parameters from a set of NN-based encoders configured for the UE; and communicate messages with the UE based on the permission and the determined NN-based encoder.

[0025] In some examples of the methods, apparatuses and nontransitory computer-readable media described herein, communication may include operations, features, components or instructions for communicating messages in incoherent transmissions, which include messages and do not have a reference signal associated with the messages for channel estimation.

[0026] In some examples of the methods, apparatuses and non-transitory computer-readable media described herein, determination may include determining a subset of a set of NN-based encoders corresponding to size parameters and determining the operation, features, components or instructions of an NN-based encoder from the subset of the NN-based encoder set based on an NN-based encoder index indicated by a license.

[0027] In some examples of the methods, apparatuses and non-transitory computer-readable media described herein, determination may include determining a subset of a set of NN-based encoders corresponding to size parameters and determining the operation, features, components or instructions of the NN-based encoders from the subset of the NN-based encoder set based on transmission parameters.

[0028] In some examples of the methods, apparatuses, and non-transitory computer-readable media described herein, transmission parameters include bandwidth indicators, modulation schemes, channel coding rates, or combinations thereof.

[0029] Some examples of the methods, apparatuses, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for sending a configuration message to a UE that indicates a set of NN-based encoders and one or more size parameters corresponding to each NN-based encoder in the set of NN-based encoders.

[0030] In some examples of the methods, apparatuses and non-transitory computer-readable media described herein, the size parameter corresponds to the TB size for the message, the number of REs for transmitting the message, the number of PRBs for transmitting the message, the number of OFDM symbols for transmitting the message, the number of coded bits for transmitting the message, or a combination thereof.

[0031] In some examples of the methods, apparatuses, and non-transitory computer-readable media described herein, communication may include operations, features, components, or instructions for modulating a message using a determined NN-based encoder and sending the modulated message to the UE.

[0032] Some examples of the methods, apparatuses, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for determining an NN-based decoder corresponding to the determined NN-based encoder. In some examples of the methods, apparatuses, and non-transitory computer-readable media described herein, communication may include operations, features, components, or instructions for receiving messages from a UE and demodulating the messages using the determined NN-based decoder.

[0033] In some examples of the methods, apparatuses and non-transitory computer-readable media described herein, licenses include dynamic uplink licenses, configured licenses, dynamic downlink licenses, SPS configurations or combinations thereof.

[0034] A method for wireless communication is described. The method may include: determining a set of segments of a resource set allocated for communication messages; determining a corresponding NN-based encoder for each segment in the set of segments based on a corresponding size parameter of each segment; and communicating messages in the resource set based on the determined corresponding NN-based encoder for each segment of the set of segments.

[0035] An apparatus for wireless communication is described. The apparatus may include a processor, a memory coupled to the processor, and instructions stored in the memory. The instructions can be executed by the processor to cause the apparatus to: determine a set of segments of a resource set allocated for communication messages; for each segment in the set of segments, determine a corresponding NNN-based encoder based on a corresponding size parameter of each segment; and communicate messages in the resource set based on the determined corresponding NNN-based encoder for each segment in the set of segments.

[0036] Another apparatus for wireless communication is described. The apparatus may include components for performing the following operations: determining a set of segments of a resource set allocated for communication messages; determining a corresponding NN-based encoder for each segment in the set of segments based on a corresponding size parameter of each segment; and communicating messages in the resource set based on the determined corresponding NN-based encoder for each segment in the set of segments.

[0037] A non-transitory computer-readable medium is described, storing code for wireless communication. The code may include processor-executable instructions for: determining a set of segments of a resource set allocated for communicating messages; determining, for each segment in the set of segments, a corresponding NN-based encoder based on a corresponding size parameter of each segment; and communicating messages in the resource set based on the determined corresponding NN-based encoder for each segment in the set of segments.

[0038] Some examples of the methods, apparatuses, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for determining whether a TB size for a message satisfies a threshold size, wherein the set of segments may be determined based on whether the TB size for the message satisfies the threshold size.

[0039] In some examples of the methods, apparatuses, and non-transitory computer-readable media described herein, communication may include operations, features, components, or instructions for modulating and transmitting a message in each segment of a set of segments using a determined, corresponding NN-based encoder for each segment in the set of segments.

[0040] Some examples of the methods, apparatuses, and non-transitory computer-readable media described herein may also include operations, features, components, or instructions for determining a corresponding NN-based decoder for each segment in a set of segments, corresponding to the determined corresponding NN-based encoder for each segment in the set of segments. In some examples of the methods, apparatuses, and non-transitory computer-readable media described herein, communication may include operations, features, components, or instructions for receiving messages and demodulating messages in each segment of the set of segments using the determined corresponding NN-based decoder for each segment in the set of segments.

[0041] In some examples of the methods, apparatuses and non-transitory computer-readable media described herein, the segment set may be determined based on one or more REs of a resource set, one or more PRBs of a resource set, one or more subbands of a resource set, one or more OFDM symbols of a resource set, one or more OFDM symbol groups of a resource set, or a combination thereof.

[0042] In some examples of the methods, apparatuses and non-transitory computer-readable media described herein, the corresponding size parameters of segments in a segment set correspond to the number of REs of that segment, the number of PRBs of that segment, the number of OFDM symbols of that segment, the number of coded bits of that segment, or a combination thereof.

[0043] In some examples of the methods, apparatuses, and non-transitory computer-readable media described herein, the first NNN-based encoder for a first segment in a set of segments may be different from the second NNN-based encoder for a second segment in the same set of segments. Attached Figure Description

[0044] Figure 1 and Figure 2 An example of a wireless communication system supporting size-based neural network (NN) selection for autoencoder-based communication is shown, according to various aspects of this disclosure.

[0045] Figure 3 An example of an autoencoder configuration supporting size-based NN selection for communication based on an autoencoder, according to various aspects of this disclosure, is shown.

[0046] Figure 4 An example of a segmented autoencoder configuration supporting size-based NN selection for communication based on autoencoders is shown, according to various aspects of this disclosure.

[0047] Figure 5 and Figure 6 An example of a processing flow supporting size-based NN selection for communication based on an autoencoder, according to various aspects of this disclosure, is shown.

[0048] Figure 7 and Figure 8 A block diagram of a device supporting size-based NN selection for communication based on an autoencoder, according to various aspects of this disclosure, is shown.

[0049] Figure 9 A block diagram of a size-based NN selection communication manager supporting communication based on autoencoders, according to various aspects of this disclosure, is shown.

[0050] Figure 10 A system is shown that includes a device for size-based NN selection for communication based on an autoencoder, according to various aspects of this disclosure.

[0051] Figure 11 and Figure 12 A block diagram of a device supporting size-based NN selection for communication based on an autoencoder, according to various aspects of this disclosure, is shown.

[0052] Figure 13 A block diagram of a size-based NN selection communication manager supporting communication based on autoencoders, according to various aspects of this disclosure, is shown.

[0053] Figure 14 A system is shown that includes a device for size-based NN selection for communication based on an autoencoder, according to various aspects of this disclosure.

[0054] Figures 15 to 19 A flowchart illustrating a method for size-based NN selection for communication based on an autoencoder, according to various aspects of this disclosure, is shown. Detailed Implementation

[0055] In some wireless communication systems, wireless devices (e.g., user equipment (UE) or base stations) can utilize autoencoder-based communication. An autoencoder can be an example of a neural network (NN) system that modulates, demodulates, or both messages. An autoencoder can include an NN-based encoder and an NN-based decoder, and the autoencoder can train the NN to compress messages for efficient transmission and decompress messages to accurately determine the compressed information, respectively. As part of the autoencoder design, the NN-based encoder and NN-based decoder can be jointly trained such that the encoder's input can be recovered as the decoder's output. Thus, a receiving device using an autoencoder with an NN-based decoder can successfully receive and decode messages from a transmitting device using an autoencoder with an NN-based encoder.

[0056] Autoencoders can be optimized or otherwise trained based on the specific message size used for communication. In some examples, wireless devices may implement multiple autoencoders to handle a range of different message sizes. However, in some cases, a wireless communication system supporting multiple autoencoders may not indicate which autoencoder (e.g., a neural network-based encoder, a neural network-based decoder) is used at the wireless device. Therefore, such a wireless device may potentially experience significant processing overhead when using an autoencoder that is unsuitable (e.g., untrained or otherwise configured) for a specific message size. For example, when employing an autoencoder (e.g., a neural network-based encoder, a neural network-based decoder) suitable for encoding or decoding relatively large messages (e.g., messages larger than a threshold size) for relatively small messages (e.g., messages larger than a threshold size), the wireless device may experience significant encoding or decoding complexity. Furthermore, if the message is being communicated in an incoherent transmission, channel estimation may not be available to support modulation or demodulation techniques for the message using a single autoencoder.

[0057] This document describes a technique for a UE and a base station to select an NN from a set of multiple NNs for autoencoder-based communication based on a size parameter for a message. In some cases, messages may be communicated in incoherent transmissions (e.g., transmissions lacking a reference signal, such as a demodulation reference signal (DMRS) for channel estimation). In other cases, the UE may be configured (e.g., configured or pre-configured by the base station) to have a set of multiple NN-based encoders. The UE may receive a permission from the base station indicating the size parameter for communicating the message, and may determine the NN-based encoder from the configured set of NN-based encoders based on this size parameter. For example, the size parameter may be the transport block (TB) size for the message or another size parameter indicated by the permission. The UE and the base station may communicate the message based on the determined NN-based encoder. For example, for uplink transmissions, the UE may use the NN-based encoder to modulate the message for transmission, and the base station may use the NN-based encoder to determine the corresponding NN-based decoder for demodulating the message. Alternatively, for downlink transmissions, the base station may use the NN-based encoder, and the UE may use the NN-based decoder. Additionally or alternatively, the UE, base station, or both may use multiple NN-based encoders, NN-based decoders, or both to communicate. For example, a wireless device (e.g., a UE or base station) may determine multiple distinct resource segments and corresponding NN-based encoders for the multiple distinct segments, based on the total set of resources allocated for communication messages.

[0058] In some examples, this configuration can map each configured NN-based encoder to a corresponding set of size parameters. In some cases, the base station can send a configuration message to the UE indicating a set of NN-based encoders and one or more size parameters corresponding to each NN-based encoder in the set. In other cases, the UE can be pre-configured with a set of NN-based encoders and one or more size parameters corresponding to each NN-based encoder in the set (e.g., in a lookup table or another format). Additionally or alternatively, a license can indicate the transmission direction for communication messages (e.g., uplink or downlink). A license can be an example of dynamic uplink license, configured license, dynamic downlink license, semi-persistent scheduling (SPS) configuration, or a combination thereof.

[0059] In some examples, multiple neural network-based encoders can be associated with specific size parameters. In some such examples, the base station can explicitly indicate the NN-based encoder index in the license, and the UE and base station can determine the NN-based encoder from multiple NN-based encoders based on the indicated index. In some other such examples, the NN-based encoder can be implicitly determined from multiple NN-based encoders based on one or more transmission parameters, where transmission parameters may include bandwidth indicators, modulation schemes, channel coding rates, or combinations thereof. Additionally or alternatively, the UE, base station, or both can determine the NN-based decoder based on a mapping from each NN-based encoder to a specific NN-based decoder, an NN-based decoder index indicated by the license, or a combination thereof.

[0060] In some examples, the UE, base station, or both can determine the implementation of resource segmentation techniques based on thresholds. For example, the device can determine to segment resources allocated for NN-based encoder selection based on whether the TB size for the message meets a threshold size. The device can determine a corresponding NN-based encoder for each segment in the set of segments based on a corresponding size parameter for each segment. For example, a first NN-based encoder determined for a first segment in the set of segments can be the same as or different from a second NN-based encoder determined for a second segment in the set of segments. Additionally or alternatively, the device can determine a corresponding NN-based decoder for each segment in the set of segments based on a corresponding size parameter for each segment, the determined corresponding NN-based encoder, or both. Each NN-based decoder can correspond to the determined corresponding NN-based encoder for each segment in the set of segments. The device can communicate messages in the allocated resources based on the determined corresponding NN-based encoder for each segment in the set of segments. For example, the device sending the message can use a determined, appropriate NN-based encoder to modulate at least a portion of the message in each resource segment to support message transmission. The device receiving the message can use a determined, appropriate NN-based decoder for each segment to demodulate the message in each resource segment.

[0061] The aspects of this disclosure are initially described in the context of wireless communication systems. Additional aspects of this disclosure are described with reference to autoencoder configuration and processing flows. Aspects of this disclosure are further illustrated and described with reference to device diagrams, system diagrams, and flowcharts relating to size-based NN selection for autoencoder-based communication.

[0062] Figure 1An example of a wireless communication system 100 supporting size-based NN selection for autoencoder-based communication is shown according to various aspects of this disclosure. The wireless communication system 100 may include one or more base stations 105, one or more UEs 115, and a core network 130. In some examples, the wireless communication system 100 may be a Long Term Evolution (LTE) network, an Advanced LTE (LTE-A) network, an LTE-A Pro network, or a New Radio (NR) network. In some examples, the wireless communication system 100 may support enhanced broadband communication, ultra-reliable (e.g., mission-critical) communication, low-latency communication, communication with low-cost and low-complexity devices, or any combination thereof.

[0063] Base stations 105 can be distributed throughout a geographical area to form a wireless communication system 100, and can be different types of devices or devices with different capabilities. Base stations 105 and UE 115 can communicate wirelessly via one or more communication links 125. Each base station 105 can provide a coverage area 110, on which UE 115 and base station 105 can establish one or more communication links 125. Coverage area 110 can be an example of a geographical area over which base station 105 and UE 115 can support signal communication according to one or more radio access technologies.

[0064] UE 115 can be distributed throughout the entire coverage area 110 of the wireless communication system 100, and each UE 115 can be stationary, mobile, or both at different times. UE 115 can be different types of devices or devices with different capabilities. Figure 1 Some example UE115s are shown in the document. The UE 115 described herein can be able to communicate with various types of devices, such as... Figure 1 Other UEs 115, base stations 105, or network equipment shown (e.g., core network nodes, relay equipment, integrated access and backhaul (IAB) nodes, or other network equipment).

[0065] Base station 105 may communicate with core network 130, or communicate with each other, or both. For example, base station 105 may interface with core network 130 via one or more backhaul links 120 (e.g., via S1, N2, N3, or other interfaces). Base station 105 may communicate with each other directly (e.g., directly between base stations 105) or indirectly (e.g., via core network 130) or both via backhaul links 120 (e.g., via X2, Xn, or other interfaces). In some examples, backhaul link 120 may be one or more radio links or include one or more radio links.

[0066] One or more base stations 105 described herein may include, or may be referred to by those skilled in the art as, base station transceiver, radio base station, access point, radio transceiver, NodeB, eNodeB (eNB), next-generation NodeB or gigabit NodeB (any of which may be referred to as gNB), home NodeB, home eNodeB or other suitable terms.

[0067] UE 115 may include or be referred to as a mobile device, wireless device, remote device, handheld device, or subscriber device, or some other suitable term, wherein "device" may also be referred to as a unit, station, terminal, client, and other examples. UE 115 may also include or be referred to as a personal electronic device, such as a cellular phone, personal digital assistant (PDA), tablet computer, laptop computer, or personal computer. In some examples, UE 115 may include or be referred to as a wireless local loop (WLL) station, Internet of Things (IoT) device, Internet of Everything (IoE) device, or machine-type communication (MTC) device, which may be implemented in various objects such as appliances, or vehicles, meters, etc.

[0068] The UE 115 described in this document can communicate with various types of devices, such as other UEs 115 that can sometimes act as relays, as well as base stations 105 and network equipment including macro eNBs or gNBs, small cell eNBs or gNBs, or relay base stations, etc. Figure 1 As shown.

[0069] UE 115 and base station 105 can wirelessly communicate with each other via one or more communication links 125 on one or more carriers. The term "carrier" can refer to a set of radio spectrum resources having a defined physical layer structure for supporting communication link 125. For example, a carrier for communication link 125 may include a portion of the radio spectrum band (e.g., a bandwidth portion (BWP)) operating according to one or more physical layer channels of a given radio access technology (e.g., LTE, LTE-A, LTE-A Pro, NR). Each physical layer channel may carry acquisition signaling (e.g., synchronization signals, system information), control signaling coordinating operations for the carrier, user data, or other signaling. Wireless communication system 100 can support communication with UE 115 using carrier aggregation or multi-carrier operation. Depending on the carrier aggregation configuration, UE 115 can be configured with multiple downlink component carriers and one or more uplink component carriers. Carrier aggregation can be used in conjunction with both frequency division duplex (FDD) and time division duplex (TDD) component carriers.

[0070] The signal waveform transmitted on a carrier can consist of multiple subcarriers (e.g., using multicarrier modulation (MCM) techniques such as Orthogonal Frequency Division Multiplexing (OFDM) or Discrete Fourier Transform Extended OFDM (DFT-S-OFDM). In a system employing MCM, a resource element can consist of one symbol period (e.g., the duration of a modulation symbol) and one subcarrier, where the symbol period and subcarrier spacing are inversely related. The number of bits carried by each resource element can depend on the modulation scheme (e.g., the order of the modulation scheme, the encoding / decoding rate of the modulation scheme, or both). Therefore, the more resource elements UE 115 receives and the higher the order of the modulation scheme, the higher the data rate UE 115 can potentially achieve. Wireless communication resources can refer to a combination of radio frequency spectrum resources, temporal resources, and spatial resources (e.g., spatial layers or beams), and using multiple spatial layers can also increase the data rate or data integrity for communication with UE 115.

[0071] One or more parameter sets (numerologies) can be supported for a carrier, where the parameter set can include subcarrier spacing (Δf) and cyclic prefix. A carrier can be divided into one or more BWPs with the same or different parameter sets. In some examples, UE 115 can be configured with multiple BWPs. In some examples, a single BWP for a carrier can be active at a given time, and UE 115 communication can be restricted to one or more active BWPs.

[0072] The time interval between base station 105 or UE 115 can be expressed as a multiple of a basic time unit, such as T. s =1 / (Δf) max ·N f The sampling period is ) seconds, where Δf max This can represent the maximum supported subcarrier spacing, and N f This can represent the maximum supported Discrete Fourier Transform (DFT) size. The time interval of a communication resource can be organized based on each radio frame having a specified duration (e.g., 10 milliseconds (ms)). Each radio frame can be identified by a System Frame Number (SFN), ranging from 0 to 1023.

[0073] Each frame may include multiple consecutively numbered subframes or time slots, and each subframe or time slot may have the same duration. In some examples, a frame may be divided (e.g., in the time domain) into subframes, and each subframe may be further divided into multiple time slots. Alternatively, each frame may include a variable number of time slots, and the number of time slots may depend on the subcarrier spacing. Each time slot may include multiple symbol periods (e.g., depending on the length of the cyclic prefix preceding each symbol period). In some wireless communication systems 100, a time slot may be further divided into multiple mini-slots containing one or more symbols. Excluding the cyclic prefix, each symbol period may contain one or more (e.g., N) symbols. f (Number) sampling periods. The duration of a symbol period can depend on the subcarrier spacing or the operating frequency band.

[0074] A subframe, time slot, mini-time slot, or symbol can be the smallest scheduling unit of the wireless communication system 100 (e.g., in the time domain) and can be referred to as a transmission time interval (TTI). In some examples, the duration of the TTI (e.g., the number of symbol periods in the TTI) can be variable. Additionally or alternatively, the smallest scheduling unit of the wireless communication system 100 can be dynamically selected (e.g., in a burst of shortened TTIs (sTTIs)).

[0075] Physical channels can be multiplexed on a carrier using various techniques. For example, one or more of Time Division Multiplexing (TDM), Frequency Division Multiplexing (FDM), or hybrid TDM-FDM techniques can be used to multiplex physical control channels and physical data channels on a downlink carrier. The control region (e.g., a control resource set (CORESET)) of the physical control channel can be defined by several symbol periods and can extend across the system bandwidth of the carrier or a subset of that bandwidth. One or more control regions (e.g., CORESETs) can be configured for a set of UEs 115. For example, one or more UEs 115 can monitor or search for control information based on one or more search space sets, and each search space set can include one or more control channel candidates in one or more aggregation levels arranged in a cascaded manner. The aggregation level of control channel candidates can refer to the number of control channel resources (e.g., control channel elements (CCEs)) associated with coded information having a control information format with a given payload size. The search space set can include a common search space set configured to send control information to multiple UEs 115 and a UE-specific search space set for sending control information to a specific UE 115.

[0076] In some examples, base station 105 may be mobile, and thus provide communication coverage for mobile geographic coverage areas 110. In some examples, different geographic coverage areas 110 associated with different technologies may overlap, but different geographic coverage areas 110 may be supported by the same base station 105. In other examples, overlapping geographic coverage areas 110 associated with different technologies may be supported by different base stations 105. Wireless communication system 100 may include, for example, a heterogeneous network, in which different types of base stations 105 use the same or different radio access technologies to provide coverage for various geographic coverage areas 110.

[0077] Wireless communication system 100 can be configured to support ultra-reliable communication or low-latency communication, or various combinations thereof. For example, wireless communication system 100 can be configured to support ultra-reliable low-latency communication (URLLC) or mission-critical communication. UE 115 can be designed to support ultra-reliable, low-latency, or mission-critical functions (e.g., mission-critical functions). Ultra-reliable communication can include private or group communication and can be supported by one or more mission-critical services such as Mission-Critical Push-to-Talk (MCPTT), Mission-Critical Video (MCVideo), or Mission-Critical Data (MCData). Support for mission-critical functions can include prioritizing services, and mission-critical services can be used for public safety or general business applications. The terms ultra-reliable, low-latency, mission-critical, and ultra-reliable low-latency are used interchangeably herein.

[0078] In some examples, UE 115 may also be able to communicate directly with other UE 115 via device-to-device (D2D) communication link 135 (e.g., using peer-to-peer (P2P) or D2D protocols). One or more UE 115s utilizing D2D communication may be within the geographic coverage area 110 of base station 105. Other UE 115s in such a group may be outside the geographic coverage area 110 of base station 105 or otherwise unable to receive transmissions from base station 105. In some examples, the group of UE 115s communicating via D2D communication may utilize a one-to-many (1:M) system in which each UE 115 transmits to every other UE 115 in the group. In some examples, base station 105 facilitates resource scheduling for D2D communication. In other cases, D2D communication is performed between UE 115s without the involvement of base station 105.

[0079] Core network 130 can provide user authentication, access authorization, tracking, Internet Protocol (IP) connectivity, and other access, routing, or mobility functions. Core network 130 can be an evolved packet core (EPC) or a 5G core (5GC), and may include at least one control plane entity (e.g., a mobility management entity (MME), access and mobility management functions (AMF)) managing access and mobility, and at least one user plane entity (e.g., a serving gateway (S-GW), packet data network (PDN) gateway (P-GW), user plane functions (UPF)) routing packets to or interconnecting with external networks. The control plane entity can manage non-access stratum (NAS) functions, such as mobility, authentication, and bearer management of UE 115 served by base station 105 associated with core network 130. User IP packets can be transmitted through the user plane entity, which can provide IP address allocation and other functions. The user plane entity can be coupled to one or more network operator IP services 150. IP services 150 may include access to the Internet, intranet(s), IP Multimedia Subsystem (IMS), or packet-switched streaming services.

[0080] Some network devices, such as base station 105, may include sub-components such as access network entity 140, which may be an example of an access node controller (ANC). Each access network entity 140 may communicate with UE 115 through one or more other access network transport entities 145, which may be referred to as a radio head, smart radio head, or transmit / receive point (TRP). Each access network transport entity 145 may include one or more antenna panels. In some configurations, the various functions of each access network entity 140 or base station 105 may be distributed across various network devices (e.g., radio heads and ANCs) or combined into a single network device (e.g., base station 105).

[0081] Wireless communication system 100 can operate using one or more frequency bands typically in the range of 300 MHz to 300 GHz. Generally, the region from 300 MHz to 3 GHz is referred to as the ultra-high frequency (UHF) region or decimeter band because the wavelength range is approximately 1 decimeter to 1 meter. UHF waves may be blocked or redirected by buildings and environmental features, but these waves can penetrate structures sufficiently to enable macrocells to provide service to UE 115 located indoors. Compared to transmissions using smaller frequencies and longer waves in the high frequency (HF) or very high frequency (VHF) portions of the spectrum below 300 MHz, UHF wave transmission can be associated with smaller antennas and shorter ranges (e.g., less than 100 km).

[0082] Wireless communication system 100 can utilize both licensed video spectrum bands and unlicensed radio spectrum bands. For example, wireless communication system 100 can employ Licensed Assisted Access (LAA), LTE Unlicensed (LTE-U) radio access technology, or NR technology in unlicensed bands such as the 5 GHz Industrial, Scientific, and Medical (ISM) band. When operating in unlicensed radio spectrum bands, devices such as base station 105 and UE 115 can employ carrier sensing for collision detection and avoidance. In some examples, operation in unlicensed bands can be based on a combination of carrier aggregation configuration and component carriers operating in licensed bands (e.g., LAA). Operation in unlicensed spectrum can include downlink transmissions, uplink transmissions, P2P transmissions, or D2D transmissions, etc.

[0083] Base station 105 or UE 115 may be equipped with multiple antennas, which can be used to employ techniques such as transmit diversity, receive diversity, multiple-input multiple-output (MIMO) communication, or beamforming. The antennas of base station 105 or UE 115 may be located within one or more antenna arrays or antenna panels that can support MIMO operation or transmit or receive beamforming. For example, one or more base station antennas or antenna arrays may be co-located at an antenna fitting such as an antenna tower. In some examples, the antennas or antenna arrays associated with base station 105 may be located in different geographical locations. Base station 105 may have an antenna array with multiple rows and columns of antenna ports, which base station 105 can use to support beamforming for communication with UE 115. Similarly, UE 115 may have one or more antenna arrays capable of supporting various MIMO or beamforming operations. Additionally or alternatively, the antenna panel may support radio frequency beamforming for signals transmitted via the antenna ports.

[0084] Beamforming (also known as spatial filtering, directional transmission, or directional reception) is a signal processing technique that can be used at a transmitting or receiving device (e.g., base station 105, UE 115) to shape or steer an antenna beam (e.g., transmit beam, receive beam) along a spatial path between the transmitting and receiving devices. Beamforming can be achieved by combining signals communicating via antenna elements of an antenna array such that some signals propagating relative to the antenna array in a particular azimuth experience constructive interference, while others experience destructive interference. Adjustments to the signals communicating via the antenna elements can include the transmitting or receiving device applying amplitude shifts, phase shifts, or both to the signals carried via the antenna elements associated with that device. The adjustments associated with each antenna element can be defined by a beamforming weight set associated with a particular azimuth (e.g., relative to the antenna array of the transmitting or receiving device, or relative to some other azimuth).

[0085] The wireless communication system 100 can be a packet-based network operating according to a layered protocol stack. In the user plane, communication at the bearer or Packet Data Convergence Protocol (PDCP) layer can be IP-based. The Radio Link Control (RLC) layer can perform packet segmentation and reassembly for communication on logical channels. The Media Access Control (MAC) layer can perform priority processing and multiplexing of logical channels to transport channels. The MAC layer can also use error detection techniques, error correction techniques, or both to support retransmission at the MAC layer to improve link efficiency. In the control plane, the Radio Resource Control (RRC) protocol layer can provide the establishment, configuration, and maintenance of the RRC connection (which supports radio bearers for user plane data) between UE 115 and base station 105 or core network 130. At the physical layer, transport channels can be mapped to physical channels.

[0086] UE 115 and base station 105 can support data retransmission to increase the likelihood of successful data reception. Hybrid Automatic Repeat Request (HARQ) feedback is a technique to increase the likelihood of correct data reception over communication link 125. HARQ can include a combination of error detection (e.g., using Cyclic Redundancy Check (CRC)), forward error correction (FEC), and retransmission (e.g., Automatic Repeat Request (ARQ)). HARQ can improve throughput at the MAC layer under poor radio conditions (e.g., low signal-to-noise ratio conditions). In some examples, the device can support same-slot HARQ feedback, where the device can provide HARQ feedback in a specific time slot for data received in a previous symbol within that time slot. In other cases, the device can provide HARQ feedback in subsequent time slots or according to some other time interval.

[0087] Some wireless communication systems 100 may support the transmission of reference signals, such as demodulation reference signals (DMRS), to support channel estimation at the receiving device. In some cases, such as in highly mobile environments and / or environments with high Doppler shift (e.g., in enhanced mobile broadband (eMBB) systems), the channel can change rapidly in the time domain. Thus, the device may transmit a large amount of reference signaling for channel estimation to accurately capture the changing channel, resulting in significant processing overhead associated with channel estimation and channel overhead associated with reference signaling. Additionally or alternatively, if the device transmits relatively small packets (e.g., below a threshold size), such as in massive machine-type communication (mMTC) systems or for sidelink communication, the reference signaling for channel estimation may dominate channel overhead (e.g., DMRS may occupy a significant portion of the transmission due to the relatively small packet size) and result in relatively poor spectral efficiency (e.g., below a threshold). Additionally or alternatively, if the channel experiences a low signal-to-noise ratio (SNR) (e.g., below a threshold), such as in mMTC systems, for sidelink communication, for Bluetooth communication, etc., channel estimation performance may be limited, resulting in relatively poor spectral efficiency (e.g., below a threshold) when implementing DMRS. In some such cases, or in other environments, wireless devices can implement incoherent transmission (e.g., transmission without DMRS). In some examples, incoherent transmission can reduce channel overhead, improve spectral efficiency, and avoid performance degradation in modulation and decoding compared to transmissions that include DMRS.

[0088] In some examples, because incoherent transmissions do not include DMRS, the wireless device receiving the incoherent transmission can estimate the channel state information (CSI) to obtain decoding metrics (e.g., maximum likelihood estimation metrics), modulation and coding schemes (MCS), constellation mapping configuration, forward error correction (FEC) coding, or any combination of these or other factors for communication on the channel. However, estimating the CSI distribution of the channel can incur large processing overhead (e.g., exceeding a threshold). Additionally or alternatively, the estimated CSI distribution may not match the actual CSI of the incoherent transmission. Furthermore, as more UEs 115 send signaling to base station 105 (e.g., in uplink multi-user single-input multiple-output (MU-SIMO) communication), as the number of constellation points increases, and as the number of time slots increases, estimating the constellation mapping for incoherent transmissions can become exponentially more complex.

[0089] In the wireless communication system 100, UE 115 and base station 105 can implement autoencoder-based communication based on size parameters for communicating messages. In some cases, as described herein, messages can be communicated in incoherent transmission. However, in other cases, messages may include DMRS or other reference signaling supporting channel estimation. UE 115 may be configured with a set of multiple NN-based encoders, each trained to handle a specific message size or size range. In some examples, UE 115 may receive a permission from base station 105 indicating size parameters for communicating messages. Based on the size parameters, UE 115 and base station 105 may each determine an NN-based encoder (e.g., the same NN-based encoder from a set of multiple NN-based encoders configured for UE 115). Additionally, UE 115 or base station 105 may determine an NN-based decoder for the message corresponding to the determined NN-based encoder. UE 115 and base station 105 can communicate messages based on the permission and on the determined NN-based encoder, the determined NN-based decoder, or both. For example, a transmitting device (e.g., UE 115 or base station 105) can use a determined NN-based encoder to transmit messages, and a receiving device (e.g., UE 115 or base station 105) can use a determined NN-based decoder to receive messages. This technique can be used for uplink communication, downlink communication, sidelink communication, backhaul communication, or any other type of wireless communication. In some examples, for different segments of allocated resources, the wireless device can use separate NN-based encoders, decoders, or both. In some such examples, the different segments of allocated resources can be determined based on the corresponding size parameters of the segments.

[0090] Figure 2 An example of a size-based NN selection wireless communication system 200 supporting autoencoder-based communication is shown according to various aspects of this disclosure. In some examples, wireless communication system 200 may implement aspects of wireless communication system 100. Wireless communication system 200 may include base station 105-a and UE 115-a, which may be examples of the corresponding devices described herein. Base station 105-a and UE 115-a may communicate via downlink communication link 205 and uplink communication link 210. Wireless communication system 200 may support autoencoder-based communication between base station 105-a and UE 115-a, which can improve spectral efficiency and signaling overhead.

[0091] In some examples, the wireless communication system 200 can support incoherent communication. For instance, wireless devices (e.g., UE 115 or base station 105) that transmit small data packets (e.g., packets smaller than a threshold) in high-mobility environments (e.g., with high Doppler shift or delay spread above a threshold) or transmit small data packets (e.g., packet sizes below a threshold) can implement incoherent communication. Additionally or alternatively, if the detected SNR value or other channel metric (e.g., signal-to-interference-plus-noise ratio (SINR)) is below a threshold, the wireless device can communicate using incoherent transmission. In such examples, the wireless device can communicate messages without a reference signal (e.g., DMRS) for channel estimation, which can reduce signaling overhead and improve spectral efficiency compared to coherent transmission (e.g., transmission with a reference signal for channel estimation). Figure 2 In some cases, base station 105-a and UE 115-a can communicate using incoherent transmission. In other cases, base station 105-a and UE 115-a can communicate messages using reference signals (e.g., DMRS) used for channel estimation.

[0092] UE 115-a and base station 105-a may implement an autoencoder for processing communications. The autoencoder may be an example of a neural network (NN) system that modulates and demodulates message 230. Additionally or alternatively, the autoencoder may encode and decode message 230. UE 115-a and base station 105-a may determine the autoencoder used for communicating message 230 (e.g., message 230-a on the uplink or message 230-b on the downlink). The autoencoder may support modulation and / or demodulation of message 230 at a sequence-based or symbol-based level. Machine learning techniques may be used to train the NN 215 of the autoencoder to determine effective data encoding (e.g., training using an unsupervised model, supervised model training, or a combination thereof). In some examples, the original equipment manufacturer (OEM) may train the NN 215 of the autoencoder offline. Additionally or alternatively, the wireless device may train the NN 215 of the autoencoder online. The autoencoder may include an NN-based encoder and an NN-based decoder. For example, NNs 215-a, 215-b, 215-c, and 215-d can each be used as an NN-based encoder, an NN-based decoder, or a combination thereof for base station 105-a. In a specific example, such as... Figure 2 As shown, NN 215-a and 215-b can each be used as NN-based encoders, and NN 215-c and 215-d can each be used as NN-based decoders for UE 115-a.

[0093] NN 215 (e.g., an NN-based encoder and a corresponding NN-based decoder) can be jointly trained to compress messages for efficient transmission and to decompress messages to accurately determine compressed data. For example, a first autoencoder may include NN215-a and NN 215-c, and NN 215-a and NN215-c can be jointly trained such that NN 215-a encodes message 230 and NN 215-c decodes message 230 (e.g., encoding based on NN 215-a). In another example, a second autoencoder may include NN 215-b and NN 215-d, and NN 215-b and NN 215-d can also be jointly trained such that NN 215-b encodes message 230 and NN 215-d decodes message 230. However, the first and second autoencoders can be trained based on different message sizes, such that the first autoencoder can handle a first range of message sizes more efficiently than the second autoencoder, and the second autoencoder can handle a second range of message sizes more efficiently than the first autoencoder. Additionally or alternatively, base station 105-a and UE 115-a can be configured with additional autoencoders for processing downlink message 230-b, uplink message 230-a, or both.

[0094] Thus, base station 105-a and UE 115-a can each be configured with one or more NNs 215, which can support different autoencoders for communication between base station 105-a and UE 115-a. However, if the device implementing multiple autoencoders does not indicate which autoencoder (e.g., NN-based encoder, NN-based decoder) is used for communication, the device receiving the encoded message can use significant processing overhead to successfully receive and decode the message (e.g., by blindly testing multiple different NN-based decoders). Furthermore, in the case where message 230 is communicated in an incoherent transmission manner, channel estimation may be inaccurate or unusable for indicating improvements in modulation or demodulation techniques, further increasing the processing complexity involved in successfully receiving message 230.

[0095] This document describes a technique for implementing autoencoder-based communication using a set of multiple neural networks 215 based on the size parameters of message 230 for UE 115-a and base station 105-a. In some cases, message 230 can be communicated in incoherent transmission. In other cases, message 230 can be communicated using a reference signal that supports channel estimation (e.g., DMRS).

[0096] In some cases, UE 115-a may be configured with a set of multiple NN-based encoders, a set of multiple NN-based decoders, or both. For example, UE 115-a may be configured with NN 215-a and NN 215-b as a set of multiple NN-based encoders. Additionally or alternatively, UE 115-a may be configured with NN 215-c and NN 215-d as a set of multiple NN-based decoders. UE 115-a may store NN 215 in memory, for example, including NN structure, NN parameters, or combinations thereof. In some cases, each NN-based encoder may correspond to an NN-based encoder index. In some examples, base station 105-a may send configuration message 220 to UE 115-a indicating a set of multiple NN-based encoders, NN-based decoders, or both. For example, configuration message 220 (e.g., Radio Resource Control (RRC) configuration message, Media Access Control (MAC) control element (CE) message, or another configuration message) may indicate the NN structure, NN parameters, NN index, or a combination thereof for a set of multiple NN-based encoders, a set of multiple NN-based decoders, or both. Additionally, configuration message 220 may indicate one or more size parameters corresponding to each NN-based encoder in the set of NN-based encoders, each NN-based decoder in the set of NN-based decoders, or both. In some other examples, UE 115-a may be pre-configured with a set of multiple NN-based encoders, one or more size parameters corresponding to each NN-based encoder in the set of NN-based encoders, a set of multiple NN-based decoders, one or more size parameters corresponding to each NN-based decoder in the set of NN-based decoders, or a combination thereof.

[0097] In some examples, UE 115-a may receive a license 225 from base station 105-a indicating size parameters for communication message 230. License 225 may also indicate the transmission direction (e.g., downlink or uplink) for communication message 230. License 225 may be an example of dynamic uplink licensing or a license configured to grant UE 115-a permission to use a set of resources for sending message 230-a on uplink communication link 210. Alternatively, license 225 may be an example of dynamic downlink licensing or a semi-persistent scheduling (SPS) physical downlink shared channel (PDSCH) configuration granting UE 115-a permission to use a set of resources for receiving message 230-b on downlink communication link 205. The size parameters indicated by the license may correspond to the transport block (TB) size for message 230, the number of resource elements (REs) allocated for message 230, the number of physical resource blocks (PRBs) allocated for message 230, the number of orthogonal frequency domain multiplexing (OFDM) symbols allocated for message 230, the number of coded bits for message 230, or a combination thereof. In some cases, the size parameters may correspond to a product of one or more size parameters (e.g., the product of the number of allocated PRBs and the number of allocated OFDM symbols, indicating the total set of resources allocated for message 230).

[0098] UE 115-a can determine an NN-based encoder from a set of multiple NN-based encoders configured at UE 115-a based on size parameters. In some cases, the NN-based encoder can be determined to encode and / or modulate message 230-a for transmission to base station 105-a. For example, UE 115-a can determine NN 215-a as the determined NN-based encoder based on the size parameters indicated in license 225. UE 115-a can use NN 215-a to modulate message 230-a and can transmit message 230-a to base station 105-a. In some cases, the determined NN-based encoder can indicate the NN-based encoder used for downlink message 230-b. For example, UE 115-a can determine NN 215-b as the NN-based encoder used for message 230-b. To receive message 230-b, UE 115-a can also determine an NN-based decoder corresponding to the NN-based encoder. For example, UE 115-a can identify NN 215-d as the NN-based decoder corresponding to NN 215-b for receiving message 230-b.

[0099] In some cases, base station 105-a can determine the NN-based encoder based on the size parameters indicated in license 225. The determined NN-based encoder for base station 105-a can be determined from a set of multiple NN-based encoders configured for UE 115-a (e.g., NNs 215-a and 215-b), or from another set of NN-based encoders configured at base station 105-a. Base station 105-a can use the determined NN-based encoder to transmit message 230-b or determine a corresponding NN-based decoder to receive message 230-a. For example, both UE 115-a and base station 105-a can determine NN 215-a as the NN-based encoder for message 230-a based on the size parameters indicated in license 225 (e.g., uplink license). Base station 105-a can also determine NN 215-c as the NN-based decoder corresponding to the determined NN-based encoder. UE 115-a can transmit message 230-a using NN 215-a as an NN-based encoder according to license 225 (e.g., use of permitted resources), and base station 10-5a can receive message 230-a using NN 215-c as an NN-based decoder according to license 225. Therefore, both UE 115-a and base station 105-a can select the NN 215 (e.g., an autoencoder) for communication based on the information included in the communication license 225.

[0100] In some cases, multiple neural network-based encoders can be associated with a single size parameter. For example, UE115-a can store Table 1 as a lookup table in memory, where Table 1 indicates the mapping between size parameters and NN 215.

[0101] Dimensions NN#1 NN#2 NN#3 NN#4 NN#5 Level 1 X X Level 2 X X X Level 3 X X

[0102] Table 1: Mapping between size parameters and neural networks (NN)

[0103] In Table 1, UE 115-a can be configured with five NNs 215 for encoding, where X indicates which size parameter corresponds to which NN 215. Each level can correspond to a range of size parameters for message 230. For example, if the size parameter is a TB size, then level 1 can correspond to a TB size below a first threshold, level 2 can correspond to a TB size greater than or equal to the first threshold and less than a second threshold, and level 3 can correspond to a TB size greater than or equal to the second threshold. As shown, in some cases, a size parameter can correspond to one or more NNs 215, and an NN 215 can correspond to one or more size parameters. If UE 115-a receives a permission 225 indicating a size parameter corresponding to multiple NNs 215 (e.g., a subset of NNs 215) in the configured set of NNs 215, then UE 115-a can select one NN 215 from the multiple NNs 215 based on an explicit or implicit indication. In some examples, base station 105-a may explicitly indicate an NN-based encoder index in license 225 (e.g., in the NN-based encoder index field of license 225) to indicate an NN-based encoder. For example, a size parameter may indicate level 1 corresponding to NN#1 and NN#2, and license 225 may indicate NN#2 as an NN-based encoder. In some other examples, UE115-a may implicitly determine the NN-based encoder based on one or more transmission parameters, such as bandwidth indicators, modulation schemes, channel coding rates, or combinations thereof. The mapping from transmission parameters(s) to the NN-based encoder may be configured at UE 115-a (e.g., configured or pre-configured by base station 105-a). For example, a size parameter may indicate level 2 corresponding to NN#2, NN#3, and NN#4, and UE115-a may determine NN#3 from this subset of NN 215 based on the bandwidth indicator used for communication. Base station 105-a can use the same set of technologies to determine the NN-based encoder (e.g., such that UE115-a and base station 105-a are coordinated to select the same NN-based encoder from a set of multiple configured NN-based encoders).

[0104] In some examples, UE 115-a can determine the NN-based decoder for message 230-b. For example, UE 115-a can determine the NN-based decoder to demodulate and / or decode message 230-b. Similarly, base station 105-a can determine the NN-based decoder to demodulate and / or decode message 230-a. In some cases, the radio devices (e.g., UE 115-a and base station 105-a) can map each NN-based encoder to an NN-based decoder. For example, UE 115-a can map NN 215-a to NN 215-c, and can map NN 215-b to NN 215-d. In this way, the NN-based decoder can be determined directly from the NN-based encoder. In some other cases, the radio devices can determine the NN-based decoder based on the NN-based decoder index indicated in license 225. For example, license 225 can indicate NN 215-d as the NN-based decoder for UE 115-a. In this way, UE 115-a can determine the autoencoder (e.g., a NN-based encoder and a corresponding NN-based decoder) used for communication.

[0105] In some examples, the wireless device can override the selection of NN 215. For example, base station 105-a can use the NN-based encoder index in license 225 to indicate NN 215-a, but UE 115-a can select NN 215-b as the NN-based encoder based on one or more selection rules. Additionally or alternatively, UE 115-a can determine NN 215-b as the NN-based encoder based on one or more selection rules (e.g., based on the size parameters indicated by license 225), and UE 115-a can alternatively select NN 215-a as the NN-based encoder based on the NN-based encoder index indicated by base station 105-a in license 225.

[0106] Using the techniques described herein, based on the size parameter for message 230, UE 115-a and base station 105-a can determine the same NN-based encoder from a set of multiple configured NN-based encoders (e.g., an NN-based encoder configured to process messages of a specific size indicated by the size parameter). UE 115-a and base station 105-a can use the described techniques for uplink transmissions on the Physical Uplink Control Channel (PUCCH), uplink transmissions on the Physical Uplink Shared Channel (PUSCH), downlink transmissions on the Physical Downlink Control Channel (PDCCH), downlink transmissions on the PDSCH, or communication on any other channel. For incoherent transmissions, the wireless device can use an NN-based decoder determined according to the NN-based encoder to demodulate the received message, thereby reducing the processing overhead and complexity involved in successfully demodulating incoherent transmissions.

[0107] Figure 3 Examples of an autoencoder configuration 300 supporting size-based NN selection for autoencoder-based communication are shown according to aspects of this disclosure. In some examples, the autoencoder configuration 300 may implement aspects of the wireless communication system 100 or 200. For example, as referenced... Figure 1 and Figure 2 The described UE 115 or base station 105 can implement autoencoder configuration 300. Autoencoder configuration 300 can use one or more autoencoders to support encoding, modulation, or both of message 320, and decoding, demodulation, or both of message 320. For example, a first device (e.g., UE 115 or base station 105) can use encoder 310 of autoencoder configuration 300 to transmit messages, and a second device (e.g., UE 115 or base station 105) can use decoder 315 of autoencoder configuration 300 to receive messages. The first device can determine which set of configured NN-based encoders 325 are used for communication, and the second device can determine which set of configured NN-based decoders 330 are used for communication. In some examples, autoencoder configuration 300 can improve the spectral efficiency and signaling overhead of communication.

[0108] In some cases, base station 105 or UE 115 can use a reference signal such as DMRS to determine the CSI measurement of channel 305. However, in the case of incoherent transmission, the CSI measurement may be unavailable to the radio device because incoherent transmission does not include a reference signal for channel estimation. Furthermore, estimating the CSI distribution (e.g., based on the channel's estimated SNR value or a Gaussian distribution or other estimated distribution) may introduce significant processing overhead at the radio device (e.g., exceeding a processing threshold). Additionally, the estimated CSI distribution may introduce inaccuracies compared to the actual CSI distribution of channel 305. UE 115 or base station 105 can implement autoencoder-based communication to improve performance on channel 305.

[0109] Autoencoder configuration 300 may be an example of a neural network (NN) system that modulates and demodulates message 320 via encoder 310 and decoder 315, respectively. Autoencoder configuration 300 may support multiple autoencoders, each of which may be jointly designed—e.g., via artificial intelligence (AI) or other machine learning techniques—to support transmission and reception. Message 320 may be input as a bit sequence to NN-based encoder 325 of encoder 310, and message 320 may be output as a bit sequence from NN-based decoder 330 of decoder 315. However, in some cases, the autoencoder may perform modulation and / or demodulation at a sequence-based or symbol-wise transmission level. The system or device may perform machine learning to train the NNs of encoder 310 and decoder 315 to determine efficient modulation techniques and decoding metrics. In some examples, the NN may reduce the complexity and signaling overhead of the communication message 320 compared to other modulation and / or demodulation techniques. NN-based encoders 325-a, 325-b, and 325-c may each correspond to the NN configured at encoder 310. Similarly, the NN-based decoders 330-a and 330-b can each correspond to the NN configured at decoder 315.

[0110] As an example, the transmitting device can select an NN-based encoder 325-a from a set of configured NN-based encoders 325 (e.g., as referenced). Figure 2The message 320 is modulated using the methods described. For example, the transmitting device can use an NN-based encoder 325-a to modulate a bit sequence representing message 320—or a portion of message 320—on the physical resources of channel 305. The transmitting device can then transmit the modulated message 320 to the receiving device via channel 305. The receiving device can select an NN-based decoder 330-a from a set of configured NN-based decoders 330 to demodulate message 320 from the physical resources and obtain an estimated bit sequence. In some examples, the NN-based decoder 330-a can determine a decoding metric for channel 305 (e.g., without DMRS for incoherent transmission). Decoder 315 can output a result message 320 based on the determined bit sequence. In some cases, a base station 105 receiving messages from multiple UEs 115 can use one or more NN-based decoders 330 to determine the estimated bit sequence for each UE 115.

[0111] In some examples, the autoencoder can be trained offline, online, or a combination thereof. For example, an OEM can train the autoencoder offline using CSI samples and can configure the trained autoencoder for one or more base stations 105, one or more UEs 115, or both. Additionally or alternatively, the wireless device can train the autoencoder online based on CSI measurements. In some examples, the wireless device can perform online training during downtime (e.g., when the processing overhead at the device is below a processing overhead threshold). Updated autoencoders can be implemented based on performance metrics. For example, if a performance metric exceeds a performance threshold, an updated autoencoder can be activated at the wireless device. In some cases, to support coordination between UE 115 and base station 105, the device training the autoencoder can send instructions of the trained autoencoder to other devices. By reducing the processing overhead used to evaluate CSI distribution, implementing an autoencoder can provide higher efficiency for incoherent communication than other modulation / demodulation techniques. Additionally or alternatively, implementing an autoencoder can improve the efficiency associated with determining decoding metrics, constellations, FEC codes, or combinations thereof.

[0112] Training an autoencoder may involve determining the weights of a neural network (NN), such as an NN-based encoder 325 and an NN-based decoder 330. For example, each NN may include an input layer, one or more hidden layers, and an output layer. In a fully connected NN with one hidden layer, each hidden layer node can receive values ​​as input from each input layer node, where each input is weighted. These NN weights may be based on a cost function adjusted during NN training. Similarly, each output layer node can receive values ​​as input from each hidden layer node, where the inputs are weighted. It should be understood that an NN may include any number of hidden layers and any number of nodes in each layer. Once the NN is trained, it can be represented using matrix operations based on the input values ​​and the determined weights. If post-deployment training (e.g., online training) is supported on the device, the device may allocate memory to store the errors and / or gradients of the inverse matrix multiplication. These errors and / or gradients may support updating the NN based on output feedback. Training a neural network (NN) can support the computation of weights (e.g., connecting input layer nodes to hidden layer nodes, connecting hidden layer nodes to other hidden layer nodes, connecting hidden layer nodes to output layer nodes, or some combination thereof) to map input patterns to desired output results. In some examples, the NN of an autoencoder can be an example of a feedforward (FF) or deep feedforward (DFF) neural network, a recurrent neural network (RNN), a long short-term memory (LSTM) neural network, or any other type of neural network.

[0113] For different message sizes, multiple autoencoders (e.g., including an NN-based encoder 325 and a corresponding NN-based decoder 330) can be trained, stored, or both differently. For example, an autoencoder may be trained for a specific size or size range of messages 320, and the autoencoder may perform relatively poorly for messages 320 of other sizes. The message size can be related to the payload size of message 320, the encoded size of message 320, or both. For example, the message size may correspond to the TB size of message 320, the number of RE, PRB, or OFDM symbols allocated for transmitting message 320, the total set of resources allocated for transmitting message 320, the number of encoded bits used for transmitting message 320, or some combination thereof. In some cases, reusing an NN such as an NN-based encoder 325 or an NN-based decoder 330 trained for small packet sizes (e.g., TB sizes below a threshold TB size) to modulate or demodulate large packets (e.g., TB sizes above a threshold TB size) may exponentially increase the complexity involved in the modulation or demodulation process, leading to inefficient processing. In some cases, for coherent transmissions, devices can use CSI measurements to aid in the modulation or demodulation of messages 320 using autoencoders, regardless of message size. For example, a device can use channel estimation results for demodulation, supporting modulation and demodulation of relatively short sequences or symbols. In some such examples, the device can use autoencoders for symbol modulation and / or demodulation. However, for incoherent transmissions, CSI measurements may not be available. In some such examples, a device can use relatively long sequences (e.g., greater than a threshold number of symbols) for modulation and / or demodulation of incoherent transmissions, and multiple autoencoders can be implemented to handle different sequence lengths.

[0114] As described herein, a device supporting autoencoder configuration 300 can determine which NN-based encoder 325 to use for communication based on the size parameters of message 320. For example, based on the indicated size parameters for communication message 320, encoder 310 can select NN-based encoder 325-c to efficiently modulate message 320. Decoder 315 can select NN-based decoder 330-b based on the indicated size parameters, the determined NN-based encoder 325-c, or both, to efficiently reconstruct the modulated message 320.

[0115] Figure 4 An example of a segmented autoencoder configuration 400 supporting size-based NN selection for autoencoder-based communication is shown according to various aspects of this disclosure. In some examples, the segmented autoencoder configuration 400 may implement aspects of wireless communication system 100 or 200, autoencoder configuration 300, or combinations thereof. For example, as referenced... Figure 1and Figure 2 The described UE 115 or base station 105 can implement segmented autoencoder configuration 400 to improve spectral efficiency and processing overhead. In some examples, UE 115, base station 105, or both can use segmented autoencoder configuration 400 to process encoding, modulation, decoding, demodulation, or some combination thereof, of relatively large messages (e.g., messages with size parameters exceeding a size parameter threshold).

[0116] In some examples, a radio device such as base station 105 or UE 115 can implement autoencoder-based communication for an allocated resource set (e.g., allocated resource set 405). In some cases, licenses (e.g., dynamic uplink licenses, configured licenses, dynamic downlink licenses, SPS configurations) can allocate resource set 405 for message communication. In some examples, the message size parameter (e.g., TB size) can exceed a size threshold. The radio device communicating the message can determine to segment the allocated resource set 405 into multiple segments 410 for encoding and decoding. In some examples, this determination can be based on the size parameter satisfying a size threshold. The radio device can determine the same or different NN-based encoders, NN-based decoders, or both for different segments 410 of the allocated resource set 405. In some examples, segmenting the allocated resource set 405 and using multiple NN-based encoders and NN-based decoders can reduce the encoding and decoding complexity of the message (e.g., compared to using a single NN-based encoder and a single NN-based decoder for the entire allocated resource set 405).

[0117] A wireless device (e.g., UE 115 or base station 105) may determine segments 410 (e.g., segments 410-a, 410-b, 410-c, and 410-d) of an allocated resource set 405 based on one or more parameters. For example, the wireless device may determine segments 410 based on the REs of the allocated resource set 405, the PRBs of the allocated resource set 405, the subband(s) of the allocated resource set 405, the OFDM symbols or OFDM symbol groups of the allocated resource set 405, the REs of segment 410, the PRBs of segment 410, the subband(s) of segment 410, the OFDM symbols or OFDM symbol groups of segment 410, or any combination thereof. Determining segments 410 may involve determining the number of segments 410, determining the size of each segment 410, determining the resource allocation for each segment 410, or a combination thereof. Segment 410 may be frequency-division multiplexed in the allocated resource set 405, time-division multiplexed in the allocated resource set 405, or some combination thereof. Additionally or alternatively, the wireless device may be configured with one or more segmentation rules to determine segment 410 based on one or more other parameters (e.g., the wireless device may be configured with the number of segments 410 to be created, the maximum segment size, the minimum segment size, or any other segmentation parameter).

[0118] In some examples, the wireless device can determine a corresponding NN-based encoder 415 for each segment 410 based on the corresponding size parameters of the segment 410. For example, the wireless device can determine an NN-based encoder 415-a for segment 410-a, an NN-based encoder 415-b for segment 410-b, an NN-based encoder 415-c for segment 410-c, and an NN-based encoder 415-d for segment 410-d. The wireless device can determine different NN-based encoders 415 for different segments 410, the same NN-based encoder 415 for different segments 410, or a combination thereof. The wireless device can determine the NN-based encoder 415-a for segment 410-a based on the corresponding size parameters of segment 410-a (e.g., the number of REs, the number of PRBs, the number of OFDM symbols, the total number of resources of segment 410-a, or some combination thereof), rather than the size parameters of the complete allocated resource set 405. Additionally or alternatively, the wireless device can determine a corresponding NN-based decoder 420 (e.g., NN-based decoders 420-a, 420-b, 420-c, and 420-d) for each segment 410 corresponding to the determined corresponding NN-based encoder 415. For example, the wireless device can determine a corresponding NN-based decoder 420-a for segment 410-a corresponding to NN-based encoder 415-a, and the wireless device can determine a corresponding NN-based decoder 420-b for segment 410-b corresponding to NN-based encoder 415-b. Similar to the NN-based encoder 415, the wireless device can determine different NN-based decoders 420 for different segments 410, the same NN-based decoder 420 for different segments 410, or a combination thereof. The NN-based encoder selection, the NN-based decoder selection, or both can be based on any of the techniques described herein (e.g., references...). Figure 2 and Figure 3 (The technology described).

[0119] A wireless device can communicate messages in an allocated resource set 405 based on a determined corresponding NN-based encoder 415 for each of the multiple segments 410 in the set. In some cases, the wireless device can use the determined corresponding NN-based encoder 415 to modulate at least a portion of the message onto resources in each segment 410 for transmission in the resource set 405. For example, the wireless device can use multiple corresponding NN-based encoders 415 to modulate multiple shorter sequences onto multiple resource segments 410, instead of using a single NN-based encoder 415 to modulate a single long sequence onto the allocated resource set 405. A receiving wireless device can receive the message through the allocated resource set 405 and can demodulate each of the multiple segments 410 in the set using a determined corresponding NN-based decoder 420. For example, the receiving device can determine multiple sequences based on multiple demodulation processes using multiple corresponding NN-based decoders 420, and can aggregate the multiple sequences to determine the message to be transmitted.

[0120] Figure 5 An example of a processing flow 500 supporting size-based NN selection for autoencoder-based communication is shown according to various aspects of this disclosure. In some examples, processing flow 500 may be implemented as described in the reference. Figure 1 and Figure 2 Aspects of the described wireless communication system 100 or 200. Processing flow 500 may include base station 105-b and UE 115-b, which may be examples of the corresponding devices described herein. Alternative examples are possible, some of which may be performed in a different order than described or not at all. In some cases, steps may include additional features not mentioned below, or further steps may be added.

[0121] In some examples, at 505, base station 105-b may send a configuration message to UE 115-b indicating multiple sets of NN-based encoders and one or more size parameters corresponding to each NN-based encoder in the multiple NN-based encoder sets. In some other examples, UE 115-b may be pre-configured with multiple sets of NN-based encoders and one or more size parameters corresponding to each NN-based encoder in the multiple NN-based encoder sets.

[0122] At point 510, base station 105-b can send a permission to UE 115-b indicating size parameters for communication messages. The permission can be an example of dynamic uplink permission, configured permission, dynamic downlink permission, SPS configuration, or a combination thereof. In some cases, the size parameters may correspond to the TB size for the message, the number of REs for transmitting the message, the number of PRBs for transmitting the message, the number of OFDM symbols for transmitting the message, the number of coded bits for transmitting the message, or a combination thereof.

[0123] At locations 515-a and 515-b, base station 105-b and UE 115-b can determine the NN-based encoder corresponding to the size parameters from a set of multiple NN-based encoders configured for UE 115-b. In some cases, base station 105-b, 115-b, or both can determine a subset of the set of multiple NN-based encoders corresponding to the size parameters, and can determine the NN-based encoder from the subset of the multiple NN-based encoder sets based on the NN-based encoder index indicated by a permission. In some other cases, base station 105-b, UE 115-b, or both can determine the NN-based encoder from the subset of the set of multiple NN-based encoders based on transmission parameters. Transmission parameters may include a bandwidth indicator, modulation scheme, channel coding rate, or a combination thereof.

[0124] At 520-a or 520-b, base station 105-b or UE 115-b can determine the NN-based decoder corresponding to the determined NN-based encoder. In some cases, the NN-based decoder can be determined based on the NN-based decoder index indicated by a license.

[0125] At 525, base station 105-b and UE 115-b can communicate with each other based on a license and the determined NN-based encoder. If the license allows resources for uplink transmission, UE 115-b can determine the NN-based encoder at 515-b, and base station 105-b can determine the NN-based decoder at 520-a. UE 115-b can use the determined NN-based encoder to modulate the message and can send the message to base station 105-b (e.g., on the uplink). Base station 105-b can receive the modulated message and can use the determined NN-based decoder to demodulate the message. If the license allows resources for downlink transmission, base station 105-b can determine the NN-based encoder at 515-a, and UE 115-b can determine the NN-based decoder at 520-b. Base station 105-b can modulate the message using a determined NN-based encoder and can transmit the message to UE 115-b (e.g., on the downlink). UE 115-b can receive the modulated message and demodulate it using a determined NN-based decoder. In some cases, the message can be communicated in incoherent transmissions, where the incoherent transmission includes the message and there is no reference signal associated with the message for channel estimation.

[0126] Figure 6 An example of a processing flow 600 supporting size-based NN selection for autoencoder-based communication is shown according to various aspects of this disclosure. In some examples, processing flow 600 may be implemented as described in the reference. Figure 1 and Figure 2 Aspects of the described wireless communication system 100 or 200. Processing flow 600 may include base station 105-c and UE 115-c, which may be examples of the corresponding devices described herein. Alternative examples are possible, some of which may be performed in a different order than described or not at all. In some cases, steps may include additional features not mentioned below, or further steps may be added.

[0127] In some cases, at 605-a and 605-b, base station 105-c and UE 115-c can determine that the size parameters (such as TB size) of the messages to be communicated between base station 105-c and UE 115-c meet the threshold size.

[0128] At locations 610-a and 610-b, base station 105-c and UE 115-c can determine a set of multiple segments of a resource set allocated for communication messages. In some cases, the set of multiple segments can be determined based on the TB size used for the message satisfying a threshold size. In some examples, the set of multiple segments can be determined based on one or more REs of the resource set, one or more PRBs of the resource set, one or more subbands of the resource set, one or more OFDM symbols of the resource set, one or more OFDM symbol groups of the resource set, or a combination thereof.

[0129] At locations 615-a and 615-b, base station 105-c and UE 115-c can determine the appropriate NN-based encoder for each segment in the set of multiple segments, based on the corresponding size parameters of each segment. In some cases, the corresponding size parameters of a segment in the set of multiple segments may correspond to the number of REs in that segment, the number of PRBs in that segment, the number of OFDM symbols in that segment, the number of coded bits in that segment, or a combination thereof. In some examples, the first NN-based encoder for a first segment in the set of multiple segments may be the same as or different from the second NN-based encoder for a second segment in the set of multiple segments.

[0130] At 620-a or 620-b, base station 105-c or UE 115-c can determine a corresponding NN-based decoder for each segment in the set of multiple segments, corresponding to the determined NN-based encoder for each segment in the set of multiple segments.

[0131] At point 625, base station 105-c and UE 115-c can communicate messages in a resource set based on a determined corresponding NN-based encoder for each segment in the set of multiple segments. If UE 115-c is sending a message to base station 105-c, UE 115-c can use the determined corresponding NN-based encoder for each segment in the set of multiple segments to modulate the message in each segment of the set of multiple segments. Base station 105-c can receive the modulated message and can use the determined corresponding NN-based decoder for each segment in the set of multiple segments to demodulate the message in each segment of the set of multiple segments. If base station 105-c is sending a message to UE 105-c, base station 105-c can use the determined corresponding NN-based encoder for each segment in the set of multiple segments to modulate the message in each segment of the set of multiple segments. UE 115-c can receive modulated messages and demodulate the messages in each segment of a set of multiple segments using a determined corresponding NN-based decoder for each segment. In some cases, messages can be communicated in incoherent transmissions, where the incoherent transmission includes the message and there is no reference signal associated with the message for channel estimation.

[0132] Figure 7 A block diagram 700 of a device 705 supporting size-based NN selection for autoencoder-based communication is shown according to various aspects of this disclosure. Device 705 may be an example of various aspects of UE 115 as described herein. Device 705 may include a receiver 710, a transmitter 715, and a communication manager 720. Device 705 may also include a processor. Each of these components may communicate with each other (e.g., via one or more buses).

[0133] Receiver 710 may provide components for receiving information associated with various information channels (e.g., control channels, data channels, and information channels related to size-based NN selection for communication based on an autoencoder), such as packets, user data, control information, or any combination thereof. Information may be transmitted to other components of device 705. Receiver 710 may utilize a single antenna or multiple antennas.

[0134] Transmitter 715 may provide components for transmitting signals generated by other components of device 705. For example, transmitter 715 may transmit information associated with various information channels (e.g., control channels, data channels, and information channels related to size-based NN selection for autoencoder-based communication), such as packets, user data, control information, or any combination thereof. In some examples, transmitter 715 may be co-located with receiver 710 in a transceiver module. Transmitter 715 may utilize a single antenna or multiple antennas.

[0135] The communication manager 720, receiver 710, transmitter 715, or various combinations thereof or various components thereof may be examples of parts for performing the various aspects of size-based NN selection for communication based on autoencoders as described herein.

[0136] In some examples, the communication manager 720, receiver 710, transmitter 715, or various combinations thereof, or components thereof, may be implemented in hardware (e.g., in a communication management circuit). The circuit may include a processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described in this disclosure.

[0137] Additionally or alternatively, in some examples, the communication manager 720, receiver 710, transmitter 715, or various combinations thereof or components thereof may be implemented using code executed by a processor (e.g., as communication management software or firmware). If implemented using code executed by a processor, the functionality of the communication manager 720, receiver 710, transmitter 715, or various combinations thereof or components thereof may be performed by a general-purpose processor, DSP, central processing unit (CPU), ASIC, FPGA, or any combination of these or other programmable logic devices.

[0138] In some examples, the communication manager 720 may be configured to use or otherwise cooperate with the receiver 710, transmitter 715, or both to perform various operations (e.g., receiving, monitoring, transmitting). For example, the communication manager 720 may receive information from the receiver 710, send information to the transmitter 715, or integrate with or combine with the receiver 710, transmitter 715, or both to receive information, send information, or perform various other operations described herein.

[0139] According to the examples disclosed herein, the communication manager 720 can support wireless communication at the UE. For example, the communication manager 720 can be configured to provide or support components for receiving a license from a base station indicating size parameters for communication messages. The communication manager 720 can be configured to provide or support components for determining the NN-based encoder corresponding to the size parameters from a set of multiple NN-based encoders configured at the UE. The communication manager 720 can be configured to provide or support components for communicating messages with the base station based on the license and the determined NN-based encoder.

[0140] Additionally or alternatively, the communication manager 720 may support wireless communication according to the examples disclosed herein. For example, the communication manager 720 may be configured to provide or support components for determining a set of multiple segments of a resource set allocated for communication messages. The communication manager 710 may be configured to provide or support components for determining a corresponding NN-based encoder for each segment in the set of multiple segments based on a corresponding size parameter of each segment. The communication manager 720 may be configured to provide or support components for communicating messages in the resource set based on the determined corresponding NN-based encoder for each segment in the set of multiple segments.

[0141] By including or configuring the communication manager 720 according to the examples described herein, device 705 (e.g., a processor that controls or is otherwise coupled to receiver 710, transmitter 715, communication manager 720, or combinations thereof) can support techniques for reducing processing resources involved in communication. For example, by determining the NN-based encoder—and correspondingly the NN-based decoder—based on size parameters, device 705 can reduce the processing complexity and overhead involved in demodulating messages. Furthermore, the techniques described herein can support reliable decoding of incoherent transmissions. Improving decoding reliability (e.g., for incoherent transmissions) can reduce the number of retransmissions performed on the channel, potentially reducing the number of times the processor increases processing power and opens processing units to handle communication. Additionally or alternatively, segmenting the allocated resources and determining appropriate NN-based encoders and decoders for different segments can reduce the processing complexity involved in modulating and demodulating signals, thereby effectively reducing processing overhead at device 705.

[0142] Figure 8A block diagram 800 of a device 805 supporting size-based NN selection for autoencoder-based communication is shown according to aspects of this disclosure. Device 805 may be an example of aspects of device 705 or UE 115 as described herein. Device 805 may include a receiver 810, a transmitter 815, and a communication manager 820. Device 805 may also include a processor. Each of these components may communicate with each other (e.g., via one or more buses).

[0143] Receiver 810 may provide components for receiving information associated with various information channels (e.g., control channels, data channels, and size-based NN selection-related information channels for communication based on autoencoders), such as packets, user data, control information, or any combination thereof. Information may be transmitted to other components of device 805. Receiver 810 may utilize a single antenna or multiple antennas.

[0144] Transmitter 815 may provide components for transmitting signals generated by other components of device 805. For example, transmitter 815 may transmit information associated with various information channels (e.g., control channels, data channels, and size-based NN selection-related information channels for autoencoder-based communication), such as packets, user data, control information, or any combination thereof. In some examples, transmitter 815 may be co-located with receiver 810 in a transceiver module. Transmitter 815 may utilize a single antenna or multiple antennas.

[0145] Device 805 or its various components may be examples of parts for performing various aspects of size-based NN selection for communication based on autoencoders as described herein. For example, communication manager 820 may include a license receiving component 825, an NN-based encoder determination component 830, a communication component 835, a segmentation component 840, or any combination thereof. Communication manager 820 may be examples of various aspects of communication manager 720 described herein. In some examples, communication manager 820 or its various components may be configured to use receiver 810, transmitter 815, or both, or otherwise cooperate with receiver 810, transmitter 815, or both to perform various operations (e.g., receiving, monitoring, transmitting). For example, communication manager 820 may receive information from receiver 810, transmit information to transmitter 815, or be integrated with receiver 810, transmitter 815, or both to receive information, transmit information, or perform various other operations as described herein.

[0146] According to the examples disclosed herein, the communication manager 820 can support wireless communication at the UE. The license receiving component 825 can be configured to provide or support components for receiving a license from the base station indicating size parameters for communication messages. The NN-based encoder determination component 830 can be configured to provide or support components for determining the NN-based encoder corresponding to the size parameters from a set of multiple NN-based encoders configured at the UE. The communication component 835 can be configured to provide or support components for communicating messages with the base station based on the license and the determined NN-based encoder.

[0147] Additionally or alternatively, the communication manager 820 may support wireless communication according to the examples disclosed herein. For example, the segmentation component 840 may be configured to provide or support components for determining a set of multiple segments of a resource set allocated for communication messages. The NN-based encoder determination component 830 may be configured to provide or support components for determining a corresponding NN-based encoder for each segment in the set of multiple segments based on a corresponding size parameter of each segment. The communication component 835 may be configured to provide or support components for communicating messages in the resource set based on the determined corresponding NN-based encoder for each segment in the set of multiple segments.

[0148] Figure 9 A block diagram 900 is shown of a communication manager 920 supporting size-based NN selection for autoencoder-based communication according to various aspects of this disclosure. The communication manager 920 may be an example of the communication manager 720, communication manager 820, or both described herein. The communication manager 920 or its various components may be examples of parts for performing the various aspects of size-based NN selection for autoencoder-based communication described herein. For example, the communication manager 920 may include a license receiving component 925, an NN-based encoder determination component 930, a communication component 935, a segmentation component 940, a configuration component 945, a modulator 950, an NN-based decoder determination component 955, a TB size thresholding component 960, a demodulator 965, or any combination thereof. Each of these components may communicate directly or indirectly with each other (e.g., via one or more buses).

[0149] According to the examples disclosed herein, the communication manager 920 can support wireless communication at the UE. The license receiving component 925 can be configured to provide or support components for receiving a license from the base station indicating size parameters for communication messages. The NN-based encoder determination component 930 can be configured to provide or support components for determining the NN-based encoder corresponding to the size parameters from a set of multiple NN-based encoders configured at the UE. The communication component 935 can be configured to provide or support components for communicating messages with the base station based on the license and the determined NN-based encoder.

[0150] In some examples, to support communication, communication component 935 may be configured to provide or support components for communicating messages in incoherent transmissions, which include messages and do not have a reference signal associated with the messages for channel estimation.

[0151] In some examples, to support determination, the NN-based encoder determination component 930 may be configured to provide or support components for determining a subset of a set of multiple NN-based encoders corresponding to size parameters. In some examples, to support determination, the NN-based encoder determination component 930 may be configured to provide or support components for determining an NN-based encoder from a subset of a set of multiple NN-based encoders based on an NN-based encoder index indicated by a license.

[0152] In some examples, to support determination, the NN-based encoder determination component 930 may be configured to provide or support components for determining a subset of a set of multiple NN-based encoders corresponding to size parameters. In some examples, to support determination, the NN-based encoder determination component 930 may be configured to provide or support components for determining an NN-based encoder from a subset of a set of multiple NN-based encoders based on transmission parameters.

[0153] In some examples, transmission parameters include bandwidth indicators, modulation schemes, channel coding rates, or combinations thereof.

[0154] In some examples, configuration component 945 may be configured to provide or support a component for receiving configuration messages from a base station that indicate a plurality of NN-based encoder sets and one or more size parameters corresponding to each NN-based encoder in the plurality of NN-based encoder sets.

[0155] In some examples, the UE is pre-configured with multiple sets of NN-based encoders and one or more size parameters corresponding to each NN-based encoder in the multiple sets of NN-based encoders.

[0156] In some examples, the size parameter corresponds to the TB size for the message, the number of REs for sending the message, the number of PRBs for sending the message, the number of OFDM symbols for sending the message, the number of coded bits for sending the message, or a combination thereof.

[0157] In some examples, to support communication, modulator 950 may be configured to provide or support components for modulating messages using a determined NN-based encoder. In some examples, to support communication, communication component 935 may be configured to provide or support components for transmitting modulated messages to a base station.

[0158] In some examples, the NN-based decoder determination component 955 may be configured to provide or support components for determining the NN-based decoder corresponding to the determined NN-based encoder. In some examples, to support communication, the communication component 935 may be configured to provide or support components for receiving messages from a base station, and the demodulator 965 may be configured to provide or support components for demodulating messages using the determined NN-based decoder.

[0159] In some examples, the NN-based decoder is determined based on the NN-based decoder index indicated by the license.

[0160] In some examples, licenses include dynamic uplink licenses, configured licenses, dynamic downlink licenses, SPS configurations, or a combination thereof.

[0161] Additionally or alternatively, the communication manager 920 may support wireless communication according to the examples disclosed herein. The segmentation component 940 may be configured to provide or support components for determining a set of multiple segments of a resource set allocated for communication messages. The NN-based encoder determination component 930 may be configured to provide or support components for determining a corresponding NN-based encoder for each segment in the set of multiple segments, based on a corresponding size parameter of each segment. The communication component 935 may be configured to provide or support components for communicating messages in the resource set based on the determined corresponding NN-based encoder for each segment in the set of multiple segments.

[0162] In some examples, the TB size threshold component 960 can be configured to provide or support a component for determining whether the TB size for a message meets a threshold size, wherein a set of multiple segments is determined based on whether the TB size for a message meets the threshold size.

[0163] In some examples, to support communication, modulator 950 may be configured to provide or support components for modulating messages in each segment of a set of multiple segments using a determined corresponding NN-based encoder for each segment. In some examples, to support communication, communication component 935 may be configured to provide or support components for transmitting modulated messages.

[0164] In some examples, the NN-based decoder determination component 955 may be configured to provide or support components for determining a corresponding NN-based decoder for each segment in a set of multiple segments, corresponding to the determined corresponding NN-based encoder for each segment in the set of multiple segments. In some examples, to support communication, the communication component 935 may be configured to provide or support components for receiving messages, and the demodulator 965 may be configured to provide or support components for demodulating messages in each segment of the set of multiple segments using the determined corresponding NN-based decoder for each segment in the set of multiple segments.

[0165] In some examples, the set of multiple segments is determined based on one or more REs of a resource set, one or more PRBs of a resource set, one or more subbands of a resource set, one or more OFDM symbols of a resource set, one or more OFDM symbol groups of a resource set, or a combination thereof.

[0166] In some examples, the corresponding size parameters of a segment in a set of multiple segments correspond to the number of REs in that segment, the number of PRBs in that segment, the number of OFDM symbols in that segment, the number of coded bits in that segment, or a combination thereof.

[0167] In some examples, the first NN-based encoder for the first segment in a set of multiple segments is different from the second NN-based encoder for the second segment in the set of multiple segments.

[0168] Figure 10A system 1000 is illustrated, including device 1005 supporting size-based NN selection for autoencoder-based communication according to various aspects of this disclosure. Device 1005 may be an example of device 705, device 805, or UE 115 as described herein, or a component including device 705, device 805, or UE 115. Device 1005 may wirelessly communicate with one or more base stations 105, UE 115, or any combination thereof. Device 1005 may include components for bidirectional voice and data communication, including components for transmitting and receiving communications, such as a communication manager 1010, I / O controller 1015, transceiver 1020, antenna 1025, memory 1030, code 1035, and processor 1040. These components may communicate electronically or be otherwise coupled (e.g., operatively, communicatively, functionally, electronically, or electrically) via one or more buses (e.g., bus 1045).

[0169] I / O controller 1015 can manage the input and output signals of device 1005. I / O controller 1015 can also manage peripheral devices not integrated into device 1005. In some cases, I / O controller 1015 can represent a physical connection or port to an external peripheral device. In some cases, I / O controller 1015 can utilize an operating system, such as... Or another well-known operating system. In some other cases, the I / O controller 1015 may represent or interact with a modem, keyboard, mouse, touchscreen, or similar device. In some cases, the I / O controller 1015 may be implemented as part of a processor such as processor 1040. In some cases, a user may interact with device 1005 via the I / O controller 1015 or via hardware components controlled by the I / O controller 1015.

[0170] In some cases, device 1005 may include a single antenna 1025. However, in other cases, device 1005 may have more than one antenna 1025, which may be able to transmit or receive multiple wireless transmissions concurrently. As described herein, transceiver 1020 may communicate bidirectionally via one or more antennas 1025, a wired or wireless link. For example, transceiver 1020 may represent a wireless transceiver and may communicate bidirectionally with another wireless transceiver. Transceiver 1020 may also include a modem for modulating packets and providing the modulated packets to one or more antennas 1025 for transmission, and demodulating packets received from one or more antennas 1025. Transceiver 1020, or transceiver 1020 and one or more antennas 1025, may be an example of transmitter 715, transmitter 815, receiver 710, receiver 810, or any combination thereof or components thereof as described herein.

[0171] Memory 1030 may include random access memory (RAM) and read-only memory (ROM). Memory 1030 may store computer-readable, computer-executable code 1035, including instructions that, when executed by processor 1040, cause device 1005 to perform the various functions described herein. Code 1035 may be stored in a non-transitory computer-readable medium such as system memory or other types of memory. In some cases, code 1035 may not be directly executable by processor 1040, but may enable a computer (e.g., when compiled and executed) to perform the functions described herein. In some cases, in addition, memory 1030 may also contain a basic I / O system (BIOS) that controls basic hardware or software operations, such as interaction with peripheral device components or devices.

[0172] Processor 1040 may include intelligent hardware devices (e.g., general-purpose processors, DSPs, CPUs, microcontrollers, ASICs, FPGAs, programmable logic devices, discrete gate or transistor logic components, discrete hardware components, or any combination thereof). In some cases, processor 1040 may be configured to operate a memory array using a memory controller. In other cases, the memory controller may be integrated into processor 1040. Processor 1040 may be configured to execute computer-readable instructions stored in memory (e.g., memory 1030) to cause device 1005 to perform various functions (e.g., functions or tasks supporting size-based NN selection for autoencoder-based communication). For example, device 1005 or components thereof may include processor 1040 and memory 1030 coupled to processor 1040, processor 1040 and memory 1030 being configured to perform the various functions described herein.

[0173] According to the examples disclosed herein, the communication manager 1010 can support wireless communication at the UE. For example, the communication manager 1010 can be configured to provide or support components for receiving a license from a base station indicating size parameters for communication messages. The communication manager 1010 can be configured to provide or support components for determining the NN-based encoder corresponding to the size parameters from a set of multiple NN-based encoders configured at the UE. The communication manager 1010 can be configured to provide or support components for communicating messages with the base station based on the license and the determined NN-based encoder.

[0174] Additionally or alternatively, the communication manager 1010 may support wireless communication according to the examples disclosed herein. For example, the communication manager 1010 may be configured to provide or support components for determining a set of multiple segments of a resource set allocated for communication messages. The communication manager 1010 may be configured to provide or support components for determining a corresponding NN-based encoder for each segment in the set of multiple segments based on a corresponding size parameter of each segment. The communication manager 1010 may be configured to provide or support components for communicating messages in the resource set based on the determined corresponding NN-based encoder for each segment in the set of multiple segments.

[0175] By including or configuring the communication manager 1010 according to the examples described herein, device 1005 can achieve one or more potential advantages. In some examples, determining the NN-based encoder based on size parameters can enable device 1005 to implement multiple NN-based encoders trained for different message sizes. Thus, dynamic selection of the NN-based encoder based on size parameters can support improved coding performance, modulation performance, or both. Similarly, selecting the NN-based decoder based on the determined NN-based encoder can support improved decoding performance, demodulation performance, or both. By determining the NN-based encoder based on size parameters, device 1005 can reduce the processing complexity and overhead involved in demodulating messages. Furthermore, the techniques described herein can support reliable decoding of incoherent transmissions, allowing for reduced channel overhead (e.g., compared to transmissions including a reference signal for channel estimation). Additionally or alternatively, segmenting the allocated resources and determining appropriate NN-based encoders and decoders for different segments can reduce the complexity involved in modulating and demodulating signals.

[0176] In some examples, the communication manager 1010 may be configured to use or otherwise cooperate with transceiver 1020, one or more antennas 1025, or any combination thereof to perform various operations (e.g., receiving, monitoring, transmitting). Although the communication manager 1010 is shown as a separate component, in some examples, one or more functions described with reference to the communication manager 1010 may be supported or performed by processor 1040, memory 1030, code 1035, or any combination thereof. For example, code 1035 may include instructions executable by processor 1040 to cause device 1005 to perform various aspects of the size-based NN selection for autoencoder-based communication described herein, or processor 1040 and memory 1030 may be configured to perform or support such operations.

[0177] Figure 11 A block diagram 1100 of a device 1105 supporting size-based NN selection for autoencoder-based communication is shown according to aspects of this disclosure. Device 1105 may be an example of aspects of base station 105 as described herein. Device 1105 may include receiver 1110, transmitter 1115, and communication manager 1120. Device 1105 may also include a processor. Each of these components may communicate with each other (e.g., via one or more buses).

[0178] Receiver 1110 may provide components for receiving information associated with various information channels (e.g., control channels, data channels, and information channels related to size-based NN selection for communication based on autoencoders), such as packets, user data, control information, or any combination thereof. Information may be passed to other components of device 1105. Receiver 1110 may utilize a single antenna or multiple antennas.

[0179] Transmitter 1115 may provide components for transmitting signals generated by other components of device 1105. For example, transmitter 1115 may transmit information associated with various information channels (e.g., control channels, data channels, and size-based NN selection-related information channels for autoencoder-based communication), such as packets, user data, control information, or any combination thereof. In some examples, transmitter 1115 may be co-located with receiver 1110 in a transceiver module. Transmitter 1115 may utilize a single antenna or multiple antennas.

[0180] The communication manager 1120, receiver 1110, transmitter 1115, or various combinations thereof or various components thereof may be examples of parts for performing the various aspects of size-based NN selection for communication based on autoencoders as described herein.

[0181] In some examples, the communication manager 1120, receiver 1110, transmitter 1115, or various combinations thereof, or components thereof, may be implemented in hardware (e.g., in a communication management circuit). The circuit may include a processor, DSP, ASIC, FPGA, or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described in this disclosure.

[0182] Additionally or alternatively, in some examples, the communication manager 1120, receiver 1110, transmitter 1115, or various combinations thereof or components thereof may be implemented as code executed by a processor (e.g., as communication management software or firmware). If implemented as processor-executed code, the functionality of the communication manager 1120, receiver 1110, transmitter 1115, or various combinations thereof or components thereof may be executed by a general-purpose processor, DSP, CPU, ASIC, FPGA, or any combination of these or other programmable logic devices.

[0183] In some examples, the communication manager 1120 may be configured to use or otherwise cooperate with the receiver 1110, transmitter 1115, or both to perform various operations (e.g., receiving, monitoring, transmitting). For example, the communication manager 1120 may receive information from the receiver 1110, send information to the transmitter 1115, or integrate with or combine with the receiver 1110, transmitter 1115, or both to receive information, send information, or perform various other operations described herein.

[0184] According to the examples disclosed herein, the communication manager 1120 can support wireless communication at a base station. For example, the communication manager 1120 can be configured to provide or support components for sending a license to the UE indicating size parameters for communication messages. The communication manager 1120 can be configured to provide or support components for determining the NN-based encoder corresponding to the size parameters from a set of multiple NN-based encoders configured for the UE. The communication manager 1120 can be configured to provide or support components for communicating messages with the UE based on the license and the determined NN-based encoder.

[0185] Additionally or alternatively, the communication manager 1120 may support wireless communication according to the examples disclosed herein. For example, the communication manager 1120 may be configured to provide or support components for determining a set of multiple segments of a resource set allocated for communication messages. The communication manager 1120 may be configured to provide or support components for determining a corresponding NN-based encoder for each segment in the set of multiple segments based on a corresponding size parameter of each segment. The communication manager 1120 may be configured to provide or support components for communicating messages in the resource set based on the determined corresponding NN-based encoder for each segment in the set of multiple segments.

[0186] Figure 12 A block diagram 1200 of a device 1205 supporting size-based neural network selection for autoencoder-based communication is shown according to aspects of this disclosure. Device 1205 may be an example of aspects of device 1105 or base station 105 as described herein. Device 1205 may include receiver 1210, transmitter 1215, and communication manager 1220. Device 1205 may also include a processor. Each of these components may communicate with each other (e.g., via one or more buses).

[0187] Receiver 1210 may provide components for receiving information associated with various information channels (e.g., control channels, data channels, and size-based NN selection-related information channels for communication based on autoencoders), such as packets, user data, control information, or any combination thereof. Information may be passed to other components of device 1205. Receiver 1210 may utilize a single antenna or multiple antennas.

[0188] Transmitter 1215 may provide components for transmitting signals generated by other components of device 1205. For example, transmitter 1215 may transmit information associated with various information channels (e.g., control channels, data channels, information channels related to size-based neural network selection for autoencoder-based communication), such as packets, user data, control information, or any combination thereof. In some examples, transmitter 1215 may be co-located with receiver 1210 in a transceiver module. Transmitter 1215 may utilize a single antenna or multiple antennas.

[0189] Device 1205 or its various components may be examples of parts for performing various aspects of size-based NN selection for communication based on an autoencoder as described herein. For example, communication manager 1220 may include a permission transmission component 1225, an NN-based encoder determination component 1230, a communication component 1235, a segmentation component 1240, or any combination thereof. Communication manager 1220 may be examples of various aspects of communication manager 1120 described herein. In some examples, communication manager 1220 or its various components may be configured to use receiver 1210, transmitter 1215, or both, or otherwise cooperate with receiver 1210, transmitter 1215, or both to perform various operations (e.g., receiving, monitoring, transmitting). For example, communication manager 1220 may receive information from receiver 1210, transmit information to transmitter 1215, or be integrated with receiver 1210, transmitter 1215, or both to receive information, transmit information, or perform various other operations as described herein.

[0190] According to the examples disclosed herein, the communication manager 1220 can support wireless communication at a base station. The license transmission component 1225 can be configured to provide or support components for transmitting a license to the UE indicating size parameters for communication messages. The NN-based encoder determination component 1230 can be configured to provide or support components for determining the NN-based encoder corresponding to the size parameters from a set of multiple NN-based encoders configured for the UE. The communication component 1235 can be configured to provide or support components for communicating messages with the UE based on the license and the determined NN-based encoder.

[0191] Additionally or alternatively, the communication manager 1220 may support wireless communication according to the examples disclosed herein. The segmentation component 1240 may be configured to provide or support components for determining a set of multiple segments of a resource set allocated for communication messages. The NN-based encoder determination component 1230 may be configured to provide or support components for determining a corresponding NN-based encoder for each segment in the set of multiple segments, based on a corresponding size parameter of each segment. The communication component 1235 may be configured to provide or support components for communicating messages in the resource set based on the determined corresponding NN-based encoder for each segment in the set of multiple segments.

[0192] Figure 13A block diagram 1300 is shown of a communication manager 1320 supporting size-based NN selection for autoencoder-based communication according to various aspects of this disclosure. The communication manager 1320 may be an example of the communication manager 1120, communication manager 1220, or both described herein. The communication manager 1320 or its various components may be examples of parts for performing the various aspects of size-based NN selection for autoencoder-based communication described herein. For example, the communication manager 1320 may include a permission transmission component 1325, an NN-based encoder determination component 1330, a communication component 1335, a segmentation component 1340, a configuration component 1345, a modulator 1350, an NN-based decoder determination component 1355, a TB size thresholding component 1360, a demodulator 1365, or any combination thereof. Each of these components may communicate directly or indirectly with each other (e.g., via one or more buses).

[0193] According to the examples disclosed herein, the communication manager 1320 can support wireless communication at a base station. The license transmission component 1325 can be configured to provide or support components for transmitting a license to the UE indicating size parameters for communication messages. The NN-based encoder determination component 1330 can be configured to provide or support components for determining the NN-based encoder corresponding to the size parameters from a set of multiple NN-based encoders configured for the UE. The communication component 1335 can be configured to provide or support components for communicating messages with the UE based on the license and the determined NN-based encoder.

[0194] In some examples, to support communication, communication component 1335 may be configured to provide or support components for communicating messages in incoherent transmissions, which include messages and do not have a reference signal associated with the messages for channel estimation.

[0195] In some examples, to support determination, the NN-based encoder determination component 1330 may be configured to provide or support components for determining a subset of a set of multiple NN-based encoders corresponding to size parameters. In some examples, to support determination, the NN-based encoder determination component 1330 may be configured to provide or support components for determining an NN-based encoder from a subset of a set of multiple NN-based encoders based on an NN-based encoder index indicated by a license.

[0196] In some examples, to support determination, the NN-based encoder determination component 1330 may be configured to provide or support components for determining a subset of a set of multiple NN-based encoders corresponding to size parameters. In some examples, to support determination, the NN-based encoder determination component 1330 may be configured to provide or support components for determining an NN-based encoder from a subset of a set of multiple NN-based encoders based on transmission parameters.

[0197] In some examples, transmission parameters include bandwidth indicators, modulation schemes, channel coding rates, or combinations thereof.

[0198] In some examples, configuration component 1345 may be configured to provide or support a component for sending a configuration message to the UE indicating a plurality of NN-based encoder sets and one or more size parameters corresponding to each NN-based encoder in the plurality of NN-based encoder sets.

[0199] In some examples, the size parameter corresponds to the TB size for the message, the number of REs for sending the message, the number of PRBs for sending the message, the number of OFDM symbols for sending the message, the number of coded bits for sending the message, or a combination thereof.

[0200] In some examples, to support communication, modulator 1350 may be configured to provide or support components for modulating messages using a determined NN-based encoder. In some examples, to support communication, communication component 1335 may be configured to provide or support components for transmitting modulated messages to the UE.

[0201] In some examples, the NN-based decoder determination component 1355 may be configured to provide or support components for determining the NN-based decoder corresponding to the determined NN-based encoder. In some examples, to support communication, the communication component 1335 may be configured to provide or support components for receiving messages from the UE, and the demodulator 1365 may be configured to provide or support components for demodulating messages using the determined NN-based decoder.

[0202] In some examples, licenses include dynamic uplink licenses, configured licenses, dynamic downlink licenses, SPS configurations, or a combination thereof.

[0203] Additionally or alternatively, the communication manager 1320 may support wireless communication according to the examples disclosed herein. The segmentation component 1340 may be configured to provide or support components for determining a set of multiple segments of a resource set allocated for communication messages. The NN-based encoder determination component 1330 may be configured to provide or support components for determining a corresponding NN-based encoder for each segment in the set of multiple segments, based on a corresponding size parameter of each segment. The communication component 1335 may be configured to provide or support components for communicating messages in the resource set based on the determined corresponding NN-based encoder for each segment in the set of multiple segments.

[0204] In some examples, the TB size threshold component 1360 can be configured to provide or support components for determining whether the TB size for a message meets a threshold size, wherein a set of multiple segments is determined based on whether the TB size for a message meets the threshold size.

[0205] In some examples, to support communication, modulator 1350 may be configured to provide or support components for modulating messages in each segment of a set of multiple segments using a determined corresponding NN-based encoder for each segment. In some examples, to support communication, communication component 1335 may be configured to provide or support components for transmitting modulated messages.

[0206] In some examples, the NN-based decoder determination component 1355 may be configured to provide or support components for determining a corresponding NN-based decoder for each segment in a set of multiple segments, corresponding to the determined corresponding NN-based encoder for each segment in the set of multiple segments. In some examples, to support communication, the communication component 1335 may be configured to provide or support components for receiving messages, and the demodulator 1365 may be configured to provide or support components for demodulating messages in each segment of the set of multiple segments using the determined corresponding NN-based decoder for each segment in the set of multiple segments.

[0207] In some examples, the set of multiple segments is determined based on one or more REs of a resource set, one or more PRBs of a resource set, one or more subbands of a resource set, one or more OFDM symbols of a resource set, one or more OFDM symbol groups of a resource set, or a combination thereof.

[0208] In some examples, the corresponding size parameters of a segment in a set of multiple segments correspond to the number of REs in that segment, the number of PRBs in that segment, the number of OFDM symbols in that segment, the number of coded bits in that segment, or a combination thereof.

[0209] In some examples, the first NN-based encoder for the first segment in a set of multiple segments is different from the second NN-based encoder for the second segment in the set of multiple segments.

[0210] Figure 14 A system 1400 is illustrated, including device 1405 supporting size-based NN selection for autoencoder-based communication according to various aspects of this disclosure. Device 1405 may be an example of device 1105, device 1205, or base station 105 as described herein, or a component including device 1105, device 1205, or base station 105. Device 1405 may wirelessly communicate with one or more base stations 105, UE 115, or any combination thereof. Device 1405 may include components for bidirectional voice and data communication, including components for transmitting and receiving communications, such as a communication manager 1410, a network communication manager 1415, a transceiver 1420, an antenna 1425, a memory 1430, a code 1435, a processor 1440, and an inter-station communication manager 1445. These components may communicate electronically or be otherwise coupled (e.g., operatively, communicatively, functionally, electronically, or electrically) via one or more buses (e.g., bus 1450).

[0211] The network communication manager 1415 can manage communication with the core network 130 (e.g., via one or more wired backhaul links). For example, the network communication manager 1415 can manage the transmission of data communication by client devices such as one or more UEs 115.

[0212] In some cases, device 1405 may include a single antenna 1425. However, in other cases, device 1405 may have more than one antenna 1425, which may be able to transmit or receive multiple wireless transmissions concurrently. As described herein, transceiver 1420 may communicate bidirectionally via one or more antennas 1425, a wired or wireless link. For example, transceiver 1420 may represent a wireless transceiver and may communicate bidirectionally with another wireless transceiver. Transceiver 1420 may also include a modem for modulating packets and providing the modulated packets to one or more antennas 1425 for transmission, and demodulating packets received from one or more antennas 1425. Transceiver 1420, or transceiver 1420 and one or more antennas 1425, may be an example of transmitter 1115, transmitter 1215, receiver 1110, receiver 1210, or any combination thereof or components thereof as described herein.

[0213] Memory 1430 may include RAM and ROM. Memory 1430 may store computer-readable, computer-executable code 1435, including instructions that, when executed by processor 1440, cause device 1405 to perform the various functions described herein. Code 1435 may be stored in a non-transitory computer-readable medium such as system memory or other types of memory. In some cases, code 1435 may not be directly executable by processor 1440, but may enable a computer (e.g., when compiled and executed) to perform the functions described herein. In some additional cases, memory 1430 may also contain a BIOS that controls basic hardware or software operations, such as interaction with peripheral device components or devices.

[0214] Processor 1440 may include intelligent hardware devices (e.g., general-purpose processors, DSPs, CPUs, microcontrollers, ASICs, FPGAs, programmable logic devices, discrete gate or transistor logic components, discrete hardware components, or any combination thereof). In some cases, processor 1440 may be configured to operate a memory array using a memory controller. In other cases, the memory controller may be integrated into processor 1440. Processor 1440 may be configured to execute computer-readable instructions stored in memory (e.g., memory 1430) to cause device 1405 to perform various functions (e.g., functions or tasks supporting size-based NN selection for autoencoder-based communication). For example, device 1405 or components thereof may include processor 1440 and memory 1430 coupled to processor 1440, processor 1440 and memory 1430 being configured to perform the various functions described herein.

[0215] Inter-site communication manager 1445 can manage communication with other base stations 105 and may include a controller or scheduler for cooperating with other base stations 105 to control communication with UE 115. For example, inter-site communication manager 1445 can coordinate the scheduling of transmissions to UE 115 for various interference mitigation techniques, such as beamforming or joint transmission. In some examples, inter-site communication manager 1445 may provide an X2 interface within LTE / LTE-A wireless communication network technology to facilitate communication between base stations 105.

[0216] According to the examples disclosed herein, the communication manager 1410 can support wireless communication at a base station. For example, the communication manager 1410 can be configured to provide or support components for sending a license to the UE indicating size parameters for communication messages. The communication manager 1410 can be configured to provide or support components for determining the NN-based encoder corresponding to the size parameters from a set of multiple NN-based encoders configured for the UE. The communication manager 1410 can be configured to provide or support components for communicating messages with the UE based on the license and the determined NN-based encoder.

[0217] Additionally or alternatively, the communication manager 1410 may support wireless communication according to the examples disclosed herein. For example, the communication manager 1410 may be configured to provide or support components for determining a set of multiple segments of a resource set allocated for communication messages. The communication manager 1410 may be configured to provide or support components for determining a corresponding NN-based encoder for each segment in the set of multiple segments based on a corresponding size parameter of each segment. The communication manager 1410 may be configured to provide or support components for communicating messages in the resource set based on the determined corresponding NN-based encoder for each segment in the set of multiple segments.

[0218] In some examples, the communication manager 1410 may be configured to use or otherwise cooperate with transceiver 1420, one or more antennas 1425, or any combination thereof to perform various operations (e.g., receiving, monitoring, transmitting). Although the communication manager 1410 is shown as a separate component, in some examples, one or more functions described with reference to the communication manager 1410 may be supported or performed by processor 1440, memory 1430, code 1435, or any combination thereof. For example, code 1435 may include instructions executable by processor 1440 to cause device 1405 to perform various aspects of the size-based NN selection for autoencoder-based communication described herein, or processor 1440 and memory 1430 may be configured to perform or support such operations.

[0219] Figure 15 A flowchart illustrating a method 1500 for size-based NN selection for autoencoder-based communication, according to various aspects of this disclosure, is shown. Operation of method 1500 can be implemented by a UE or its components as described herein. For example, operation of method 1500 can be achieved by, as referenced... Figures 1 to 10The UE 115 described is executed. In some examples, the UE may execute a set of instructions to control the functional elements of the UE to perform the described functions. Additionally or alternatively, the UE may use dedicated hardware to perform aspects of the described functions.

[0220] At point 1505, the method may include receiving permission from the base station indicating size parameters for a communication message. The operation of point 1505 can be performed according to the method described herein. In some examples, aspects of the operation of point 1505 may be derived from, as referenced... Figure 9 The described license receiving component 925 is executed.

[0221] At 1510, the method may include determining the NN-based encoder corresponding to the size parameter from a set of multiple NN-based encoders configured at the UE. The operation of 1510 can be performed according to the method described herein. In some examples, aspects of the operation of 1510 may be derived from, as referenced... Figure 9 The described NN-based encoder determines the execution of component 930.

[0222] At point 1515, the method may include communicating messages with the base station based on permission and the determined NN encoder. The operation of point 1515 can be performed according to the method described herein. In some examples, aspects of the operation of point 1515 may be derived from, as referenced... Figure 9 The described communication component 935 is executed.

[0223] Figure 16 A flowchart illustrating a size-based NN selection method 1600 for autoencoder-based communication, according to various aspects of this disclosure, is shown. Operation of method 1600 can be implemented by a UE or its components as described herein. For example, operation of method 1600 can be achieved by, as described in reference... Figures 1 to 10 The UE 115 described is executed. In some examples, the UE may execute a set of instructions to control the functional elements of the UE to perform the described functions. Additionally or alternatively, the UE may use dedicated hardware to perform aspects of the described functions.

[0224] At 1605, the method may include receiving permission from the base station indicating size parameters for a communication message. The operation of 1605 can be performed according to the method described herein. In some examples, aspects of the operation of 1605 may be derived from, as referenced... Figure 9 The described license receiving component 925 is executed.

[0225] At 1610, the method may include determining the NN-based encoder corresponding to the size parameter from a set of multiple NN-based encoders configured at the UE. The operation of 1610 can be performed according to the method described herein. In some examples, aspects of the operation of 1610 may be derived from, as referenced... Figure 9 The described NN-based encoder determines the execution of component 930.

[0226] At 1615, the method may include communicating messages with the base station based on a permission and a determined NN-based encoder. In some examples, communication may involve modulating the message using the determined NN-based encoder and sending the modulated message to the base station. The operation of 1615 can be performed according to the method described herein. In some examples, aspects of the operation of 1615 may be derived from, as referenced... Figure 9 The described communication component 935, modulator 950, or any combination thereof are executed.

[0227] Figure 17 A flowchart illustrating a method 1700 for size-based NN selection for autoencoder-based communication, according to various aspects of this disclosure, is shown. Operation of method 1700 can be implemented by a UE or its components as described herein. For example, operation of method 1700 can be achieved by, as described in reference... Figures 1 to 10 The UE 115 described is executed. In some examples, the UE may execute a set of instructions to control the functional elements of the UE to perform the described functions. Additionally or alternatively, the UE may use dedicated hardware to perform aspects of the described functions.

[0228] At 1705, the method may include receiving permission from the base station indicating size parameters for a communication message. The operation of 1705 can be performed according to the method described herein. In some examples, aspects of the operation of 1705 may be derived from, as referenced... Figure 9 The described license receiving component 925 is executed.

[0229] At 1710, the method may include determining the NN-based encoder corresponding to the size parameter from a set of multiple NN-based encoders configured at the UE. The operation of 1710 can be performed according to the method described herein. In some examples, aspects of the operation of 1710 may be derived from, as referenced... Figure 9 The described NN-based encoder determines the execution of component 930.

[0230] At 1715, the method may include determining a neural network-based decoder corresponding to the determined neural network-based encoder. The operation at 1715 can be performed according to the method described herein. In some examples, aspects of the operation at 1715 may be derived from, as referenced... Figure 9The described NN-based decoder determines the execution of component 955.

[0231] At 1720, the method may include communicating messages with the base station according to permission and based on a determined NN encoder. In some examples, communication may involve receiving messages from the base station and demodulating the messages using the determined NN-based decoder. The operation of 1720 can be performed according to the method described herein. In some examples, aspects of the operation of 1720 may be derived from, as referenced... Figure 9 The described communication component 935, demodulator 965, or any combination thereof are executed.

[0232] Figure 18 A flowchart illustrating a size-based NN selection method 1800 for autoencoder-based communication, according to various aspects of this disclosure, is shown. Operation of method 1800 can be implemented by a base station or its components as described herein. For example, operation of method 1800 can be implemented by [reference needed]. Figures 1 to 5 as well as Figures 11 to 14 The described base station 105 performs the following: In some examples, the base station may execute a set of instructions to control the functional elements of the base station to perform the described functions. Additionally or alternatively, the base station may use dedicated hardware to perform aspects of the described functions.

[0233] At 1805, the method may include sending permission to the UE indicating size parameters for communication messages. The operation at 1805 can be performed according to the method described herein. In some examples, aspects of the operation at 1805 may be derived from, as referenced... Figure 13 The license sending component 1325 described is executed.

[0234] At 1810, the method may include determining the NN-based encoder corresponding to the size parameter from a set of multiple NN-based encoders configured for the UE. The operation of 1810 can be performed according to the method described herein. In some examples, aspects of the operation of 1810 may be derived from, as referenced... Figure 13 The described NN-based encoder determines the execution of component 1330.

[0235] At point 1815, the method may include communicating messages with the UE based on permission and the determined NN encoder. The operation of point 1815 can be performed according to the method described herein. In some examples, aspects of the operation of point 1815 may be derived from, as referenced... Figure 13 The described communication component 1335 is executed.

[0236] Figure 19A flowchart illustrating a size-based NN selection method 1900 for autoencoder-based communication, according to various aspects of this disclosure, is shown. Operation of method 1900 can be implemented by a UE or its components as described herein, or by a base station or its components. For example, operation of method 1900 can be implemented by, as described in reference... Figures 1 to 10 The UE described or referenced Figures 1 to 5 as well as Figures 11 to 14 The described base station 105 performs the functions described. In some examples, the UE or base station may execute a set of instructions to control the functional elements of the UE or base station to perform the described functions. Additionally or alternatively, the UE or base station may use dedicated hardware to perform aspects of the described functions.

[0237] At 1905, the method may include determining a set of multiple segments of resources allocated for communicating messages. The operation at 1905 can be performed according to the method described herein. In some examples, aspects of the operation at 1905 may be derived from, as referenced... Figures 9 to 13 The segmented component 940 or segmented component 1340 described is executed.

[0238] At 1910, the method may include determining the appropriate NN-based encoder for each segment in a set of multiple segments based on the corresponding size parameter of each segment. The operation at 1910 can be performed according to the method described herein. In some examples, aspects of the operation at 1910 may be derived from, as referenced... Figures 9 to 13 The described NN-based encoder determination component 930 or NN-based encoder determination component 1330 is executed.

[0239] At 1915, the method may include communicating messages in a resource set based on a determined corresponding NN-based encoder for each segment in a set of multiple segments. The operation at 1915 can be performed according to the method described herein. In some examples, aspects of the operation at 1915 may be derived from, as referenced... Figures 9 to 13 The described communication component 935 or communication component 1335 is executed.

[0240] It should be noted that the methods described in this paper describe possible implementations, and the operations and steps can be rearranged or otherwise modified, and other implementations are possible. Furthermore, aspects from two or more methods can be combined.

[0241] While aspects of LTE, LTE-A, LTE-A Pro, or NR systems may be described for illustrative purposes, and the terms LTE, LTE-A, LTE-A Pro, or NR may be used in most of the description, the techniques described herein are applicable beyond LTE, LTE-A, LTE-A Pro, or NR networks. For example, the techniques described are applicable to a variety of other wireless communication systems, such as Ultra Mobile Broadband (UMB), IEEE 802.11 (Wi-Fi), IEEE 802.16 (WiMAX), IEEE 802.20, Flash-OFDM, and other systems and radio technologies not explicitly mentioned herein.

[0242] The information and signals described herein can be represented using any of a variety of different technologies and techniques. For example, data, instructions, commands, information, signals, bits, symbols, and chips that may be mentioned throughout this specification can be represented by voltage, current, electromagnetic waves, magnetic fields or magnetic particles, light fields or light particles, or any combination thereof.

[0243] The various illustrative blocks and components described herein can be implemented or executed using a general-purpose processor, DSP, ASIC, CPU, FPGA, or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general-purpose processor may be a microprocessor, but alternatively, the processor may be any processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices (e.g., a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors combined with a DSP core, or any other such configuration).

[0244] The functions described herein can be implemented in hardware, software executed by a processor, firmware, or any combination thereof. If implemented in software executed by a processor, the functions can be stored on or transmitted via a computer-readable medium as one or more instructions or code. Other examples and implementations are within the scope of this disclosure and the appended claims. For example, due to the nature of software, the functions described herein can be implemented using software executed by a processor, hardware, firmware, hardwired, or any combination thereof. Features implementing the functions can also be physically located in various locations, including being distributed such that portions of the functions are implemented in different physical locations.

[0245] Computer-readable media includes both non-transitory computer storage media and communication media, with communication media encompassing any medium that facilitates the transfer of a computer program from one place to another. Non-transitory storage media can be any available medium accessible by a general-purpose or special-purpose computer. By way of example, and not limitation, non-transitory computer-readable media may include RAM, ROM, electrically erasable programmable ROM (EEPROM), flash memory, compressed optical disc (CD) ROM or other optical disc storage, disk storage or other magnetic storage devices, or any other non-transitory medium that can be used to carry or store required program code components in the form of instructions or data structures and is accessible by a general-purpose or special-purpose computer or a general-purpose or special-purpose processor. Furthermore, any connection is appropriately referred to as computer-readable media. For example, if software is transmitted from a website, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, then coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of computer-readable media. As used herein, disks and optical discs include CDs, laser discs, optical discs, DVDs, floppy disks, and Blu-ray discs, where disks typically reproduce data magnetically, while optical discs reproduce data optically using lasers. Combinations of these are also included within the scope of computer-readable media.

[0246] As used herein, including in the claims, the use of "or" in the list of items (e.g., a list of items ending with phrases such as "at least one of..." or "one or more of...") indicates an inclusive list, such that 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). Furthermore, as used herein, the phrase "based on" should not be construed as a reference to a set of closing conditions. For example, an example step described as "based on condition A" may be based on both condition A and condition B without departing from the scope of this disclosure. In other words, as used herein, the phrase "based on" should be interpreted in the same manner as the phrase "at least partially based on".

[0247] In the accompanying drawings, similar components or features may have the same reference numerals. Furthermore, various components of the same type can be distinguished by a dash following the reference numeral and a second reference numeral used to differentiate between similar components. If only the first reference numeral is used in the specification, this specification applies to any similar component having the same first reference numeral, without regard to the second reference numeral or other subsequent reference numerals.

[0248] The description herein, illustrated with reference to the accompanying drawings, describes exemplary configurations and does not represent all examples that can be implemented or that are within the scope of the claims. The term "example" as used herein means "serving as an example, instance, or illustration," and not "preferred" or "superior to other examples." For the purpose of providing an understanding of the techniques described, the detailed description includes specific details. However, these techniques can be practiced without these specific details. In some instances, well-known structures and devices are shown in block diagram form to avoid obscuring the concepts of the described examples.

[0249] The description herein is provided to enable those skilled in the art to make or use this disclosure. Various modifications to this disclosure will be apparent to those skilled in the art, and the general principles defined herein may be applied to other variations without departing from the scope of this disclosure. Therefore, this disclosure is not limited to the examples and designs described herein, but is to be given the broadest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for performing wireless communication at a user equipment (UE), comprising: Receive permission from network devices to indicate size parameters used for communication messages; The neural network-based encoder corresponding to the size parameter is determined from a plurality of neural network-based encoders configured at the UE; as well as The message is communicated with the network device in accordance with the permission and at least in part based on the determined neural network-based encoder.

2. The method according to claim 1, wherein, The communication includes: The message is communicated in an incoherent transmission, which includes the message and has no associated reference signal for channel estimation.

3. The method according to claim 1, wherein, The determination includes: Determine a subset of the plurality of neural network-based encoders corresponding to the size parameters; and The neural network-based encoder is determined from the subset of the plurality of neural network-based encoders, at least in part, based on the neural network-based encoder index indicated by the license.

4. The method according to claim 1, wherein, The determination includes: Determine a subset of the plurality of neural network-based encoders corresponding to the size parameters; and The neural network-based encoder is determined from the subset of the plurality of neural network-based encoders, at least in part, based on transmission parameters.

5. The method according to claim 4, wherein, The transmission parameters include bandwidth indicator, modulation scheme, channel coding rate, or a combination thereof.

6. The method according to claim 1, further comprising: The network device receives a configuration message indicating the plurality of neural network-based encoders and one or more size parameters corresponding to each of the plurality of neural network-based encoders.

7. The method according to claim 1, wherein, The UE is pre-configured with the plurality of neural network-based encoders and one or more size parameters corresponding to each of the plurality of neural network-based encoders.

8. The method according to claim 1, wherein, The size parameters correspond to the transport block size for the message, the number of resource elements for sending the message, the number of physical resource blocks for sending the message, the number of orthogonal frequency domain multiplexing symbols for sending the message, the number of coded bits for sending the message, or a combination thereof.

9. The method according to claim 1, wherein, The communication includes: The message is modulated using the determined neural network-based encoder; and Send a modulated message to the network device.

10. The method according to claim 1, further comprising: Determine a neural network-based decoder corresponding to the determined neural network-based encoder, wherein the communication includes: Receive the message from the network device; and The message is demodulated using the determined neural network-based decoder.

11. The method according to claim 10, wherein, The neural network-based decoder is determined at least in part based on the neural network-based decoder index indicated by the license.

12. The method according to claim 1, wherein, The licenses include dynamic uplink licenses, configured licenses, dynamic downlink licenses, semi-persistent scheduling configurations, or combinations thereof.

13. A method for wireless communication at a network device, comprising: Send permission to the user equipment (UE) indicating the size parameters used for communication messages; The neural network-based encoder corresponding to the size parameter is determined from a plurality of neural network-based encoders configured for the UE; as well as The message is communicated with the UE in accordance with the permission and at least in part based on the determined neural network-based encoder.

14. The method according to claim 13, wherein, The communication includes: The message is communicated in an incoherent transmission, which includes the message and has no associated reference signal for channel estimation.

15. The method according to claim 13, wherein, The determination includes: Determine a subset of the plurality of neural network-based encoders corresponding to the size parameters; and The neural network-based encoder is determined from the subset of the plurality of neural network-based encoders, at least in part, based on the neural network-based encoder index indicated by the license.

16. The method according to claim 13, wherein, The determination includes: Determine a subset of the plurality of neural network-based encoders corresponding to the size parameters; and The neural network-based encoder is determined from the subset of the plurality of neural network-based encoders, at least in part, based on transmission parameters.

17. The method according to claim 16, wherein, The transmission parameters include bandwidth indicator, modulation scheme, channel coding rate, or a combination thereof.

18. The method of claim 13, further comprising: A configuration message is sent to the UE indicating the plurality of neural network-based encoders and one or more size parameters corresponding to each of the plurality of neural network-based encoders.

19. The method according to claim 13, wherein, The size parameters correspond to the transport block size for the message, the number of resource elements for sending the message, the number of physical resource blocks for sending the message, the number of orthogonal frequency domain multiplexing symbols for sending the message, the number of coded bits for sending the message, or a combination thereof.

20. The method according to claim 13, wherein, The communication includes: The message is modulated using the determined neural network-based encoder; and The modulated message is sent to the UE.

21. The method of claim 13, further comprising: Determine a neural network-based decoder corresponding to the determined neural network-based encoder, wherein the communication includes: Receive the message from the UE; and The message is demodulated using the determined neural network-based decoder.

22. The method according to claim 13, wherein, The licenses include dynamic uplink licenses, configured licenses, dynamic downlink licenses, semi-persistent scheduling configurations, or combinations thereof.

23. A method for wireless communication, comprising: Identify multiple segments of the resource set allocated for communication messages; For each of the plurality of segments, the corresponding neural network-based encoder is determined at least in part based on the corresponding size parameters of each segment; as well as The messages are communicated in the resource set based at least in part on a corresponding neural network-based encoder determined for each of the plurality of segments.

24. The method of claim 23, further comprising: The transport block size for the message is determined to satisfy a threshold size, wherein the plurality of segments are determined at least in part based on the transport block size for the message satisfying the threshold size.

25. The method according to claim 23, wherein, The communication includes: The message is modulated in each of the plurality of segments using a corresponding neural network-based encoder determined for each of the segments; and Send the modulated message.

26. The method of claim 23, further comprising: For each of the plurality of segments, a corresponding neural network-based decoder is determined corresponding to the determined corresponding neural network-based encoder for each of the plurality of segments, wherein the communication includes: Receive the message; and The message is demodulated in each of the plurality of segments using a corresponding neural network-based decoder determined for each of the plurality of segments.

27. The method according to claim 23, wherein, The plurality of segments are determined at least in part based on one or more resource elements of the resource set, one or more physical resource blocks of the resource set, one or more subbands of the resource set, one or more orthogonal frequency domain multiplexing symbols of the resource set, one or more groups of orthogonal frequency domain multiplexing symbols of the resource set, or a combination thereof.

28. The method according to claim 23, wherein, The corresponding size parameters of the segments in the plurality of segments correspond to the number of resource elements of the segment, the number of physical resource blocks of the segment, the number of orthogonal frequency domain multiplexing symbols of the segment, the number of coded bits of the segment, or a combination thereof.

29. The method according to claim 23, wherein, The first neural network-based encoder used for the first segment of the plurality of segments is different from the second neural network-based encoder used for the second segment of the plurality of segments.

30. An apparatus for performing wireless communication at a user equipment (UE), comprising: processor; Memory, coupled to the processor; as well as Instructions stored in the memory and executable by the processor, to cause the device to: Receive permission from network devices to indicate size parameters used for communication messages; The neural network-based encoder corresponding to the size parameter is determined from a plurality of neural network-based encoders configured at the UE; as well as The message is communicated with the network device in accordance with the permission and at least in part based on the determined neural network-based encoder.

31. The apparatus according to claim 30, wherein, The instructions for communication can be executed by the processor to enable the device to: The message is communicated in an incoherent transmission, which includes the message and has no associated reference signal for channel estimation.

32. The apparatus according to claim 30, wherein, The instructions for determining can be executed by the processor to enable the device to: Determine a subset of the plurality of neural network-based encoders corresponding to the size parameters; as well as The neural network-based encoder is determined from the subset of the plurality of neural network-based encoders, at least in part, based on the neural network-based encoder index indicated by the license.

33. The apparatus according to claim 30, wherein, The instructions for determining can be executed by the processor to enable the device to: Determine a subset of the plurality of neural network-based encoders corresponding to the size parameters; as well as The neural network-based encoder is determined from the subset of the plurality of neural network-based encoders, at least in part, based on transmission parameters.

34. The apparatus according to claim 33, wherein, The transmission parameters include bandwidth indicator, modulation scheme, channel coding rate, or a combination thereof.

35. The apparatus according to claim 30, wherein, The instructions can also be executed by the processor to make the device: The network device receives a configuration message indicating the plurality of neural network-based encoders and one or more size parameters corresponding to each of the plurality of neural network-based encoders.

36. The apparatus according to claim 30, wherein, The UE is pre-configured with the plurality of neural network-based encoders and one or more size parameters corresponding to each of the plurality of neural network-based encoders.

37. The apparatus according to claim 30, wherein, The size parameters correspond to the transport block size for the message, the number of resource elements for sending the message, the number of physical resource blocks for sending the message, the number of orthogonal frequency domain multiplexing symbols for sending the message, the number of coded bits for sending the message, or a combination thereof.

38. The apparatus according to claim 30, wherein, The instructions for communication can be executed by the processor to enable the device to: The message is modulated using the determined neural network-based encoder; as well as Send a modulated message to the network device.

39. The apparatus according to claim 30, wherein, The instructions can also be executed by the processor to make the device: Determine a neural network-based decoder corresponding to the determined neural network-based encoder, wherein the instructions for communication can be executed by the processor to cause the device to: Receive the message from the network device; and The message is demodulated using the determined neural network-based decoder.

40. The apparatus according to claim 39, wherein, The neural network-based decoder is determined at least in part based on the neural network-based decoder index indicated by the license.

41. The apparatus according to claim 30, wherein, The licenses include dynamic uplink licenses, configured licenses, dynamic downlink licenses, semi-persistent scheduling configurations, or combinations thereof.

42. An apparatus for wireless communication at a network device, comprising: processor; Memory, coupled to the processor; as well as Instructions stored in the memory and executable by the processor, to cause the device to: Send permission to the user equipment (UE) indicating the size parameters used for communication messages; The neural network-based encoder corresponding to the size parameter is determined from a plurality of neural network-based encoders configured for the UE; as well as The message is communicated with the UE in accordance with the permission and at least in part based on the determined neural network-based encoder.

43. The apparatus according to claim 42, wherein, The instructions for communication can be executed by the processor to enable the device to: The message is communicated in an incoherent transmission, which includes the message and has no associated reference signal for channel estimation.

44. The apparatus according to claim 42, wherein, The instructions for determining can be executed by the processor to enable the device to: Determine a subset of the plurality of neural network-based encoders corresponding to the size parameters; as well as The neural network-based encoder is determined from the subset of the plurality of neural network-based encoders, at least in part, based on the neural network-based encoder index indicated by the license.

45. The apparatus according to claim 42, wherein, The instructions for determining can be executed by the processor to enable the device to: Determine a subset of the plurality of neural network-based encoders corresponding to the size parameters; as well as The neural network-based encoder is determined from the subset of the plurality of neural network-based encoders, at least in part, based on transmission parameters.

46. ​​The apparatus according to claim 45, wherein, The transmission parameters include bandwidth indicator, modulation scheme, channel coding rate, or a combination thereof.

47. The apparatus according to claim 42, wherein, The instructions can also be executed by the processor to make the device: A configuration message is sent to the UE indicating the plurality of neural network-based encoders and one or more size parameters corresponding to each of the plurality of neural network-based encoders.

48. The apparatus according to claim 42, wherein, The size parameters correspond to the transport block size for the message, the number of resource elements for sending the message, the number of physical resource blocks for sending the message, the number of orthogonal frequency domain multiplexing symbols for sending the message, the number of coded bits for sending the message, or a combination thereof.

49. The apparatus according to claim 42, wherein, The instructions for communication can be executed by the processor to enable the device to: The message is modulated using the determined neural network-based encoder; as well as The modulated message is sent to the UE.

50. The apparatus according to claim 42, wherein, The instructions can also be executed by the processor to make the device: Determine a neural network-based decoder corresponding to the determined neural network-based encoder, wherein the instructions for communication can be executed by the processor to cause the device to: Receive the message from the UE; and The message is demodulated using the determined neural network-based decoder.

51. The apparatus according to claim 42, wherein, The licenses include dynamic uplink licenses, configured licenses, dynamic downlink licenses, semi-persistent scheduling configurations, or combinations thereof.

52. An apparatus for wireless communication, comprising: processor; Memory, coupled to the processor; as well as Instructions stored in the memory and executable by the processor, to cause the device to: Identify multiple segments of the resource set allocated for communication messages; For each of the plurality of segments, the corresponding neural network-based encoder is determined at least in part based on the corresponding size parameters of each segment; as well as The messages are communicated in the resource set based at least in part on a corresponding neural network-based encoder determined for each of the plurality of segments.

53. The apparatus according to claim 52, wherein, The instructions can also be executed by the processor to make the device: The transport block size for the message is determined to satisfy a threshold size, wherein the plurality of segments are determined at least in part based on the transport block size for the message satisfying the threshold size.

54. The apparatus according to claim 52, wherein, The instructions for communication can be executed by the processor to enable the device to: The message is modulated in each of the plurality of segments using a corresponding neural network-based encoder determined for each of the plurality of segments; as well as Send the modulated message.

55. The apparatus according to claim 52, wherein, The instructions can also be executed by the processor to make the device: For each of the plurality of segments, a corresponding neural network-based decoder is determined corresponding to the determined corresponding neural network-based encoder for each of the plurality of segments, wherein the instructions for communication can be executed by the processor to cause the device to: Receive the message; and The message is demodulated in each of the plurality of segments using a determined corresponding neural network-based decoder for each of the plurality of segments.

56. The apparatus according to claim 52, wherein, The plurality of segments are determined at least in part based on one or more resource elements of the resource set, one or more physical resource blocks of the resource set, one or more subbands of the resource set, one or more orthogonal frequency domain multiplexing symbols of the resource set, one or more groups of orthogonal frequency domain multiplexing symbols of the resource set, or a combination thereof.

57. The apparatus according to claim 52, wherein, The corresponding size parameters of the segments in the plurality of segments correspond to the number of resource elements of the segment, the number of physical resource blocks of the segment, the number of orthogonal frequency domain multiplexing symbols of the segment, the number of coded bits of the segment, or a combination thereof.

58. The apparatus according to claim 52, wherein, The first neural network-based encoder used for the first segment of the plurality of segments is different from the second neural network-based encoder used for the second segment of the plurality of segments.

59. An apparatus for performing wireless communication at a user equipment (UE), comprising: A component for receiving permission from a network device that indicates dimensional parameters for use in communication messages; A component for determining the neural network-based encoder corresponding to the size parameter from a plurality of neural network-based encoders configured at the UE; as well as A component for communicating the message with the network device in accordance with the permission and at least in part based on the determined neural network-based encoder.

60. The apparatus according to claim 59, wherein, The communication component includes: Components for communicating the message in an incoherent transmission, the incoherent transmission including the message and without a reference signal associated with the message for channel estimation.

61. The apparatus according to claim 59, wherein, The components used for determining include: Components for determining subsets of the plurality of neural network-based encoders corresponding to the size parameters; and A component for determining the neural network-based encoder from the subset of the plurality of neural network-based encoders, at least in part, based on the neural network-based encoder index indicated by the license.

62. The apparatus according to claim 59, wherein, The components used for determining include: Components for determining subsets of the plurality of neural network-based encoders corresponding to the size parameters; and Components for determining the neural network-based encoder from the subset of the plurality of neural network-based encoders based at least in part on transmission parameters.

63. The apparatus according to claim 62, wherein, The transmission parameters include bandwidth indicator, modulation scheme, channel coding rate, or a combination thereof.

64. The apparatus of claim 59, further comprising: A component for receiving configuration messages from the network device that indicate the plurality of neural network-based encoders and one or more size parameters corresponding to each of the plurality of neural network-based encoders.

65. The apparatus according to claim 59, wherein, The UE is pre-configured with the plurality of neural network-based encoders and one or more size parameters corresponding to each of the plurality of neural network-based encoders.

66. The apparatus according to claim 59, wherein, The size parameters correspond to the transport block size for the message, the number of resource elements for sending the message, the number of physical resource blocks for sending the message, the number of orthogonal frequency domain multiplexing symbols for sending the message, the number of coded bits for sending the message, or a combination thereof.

67. The apparatus according to claim 59, wherein, The communication component includes: Components for modulating the message using the determined neural network-based encoder; and A component used to send modulated messages to the network device.

68. The apparatus of claim 59, further comprising: Components for determining a neural network-based decoder corresponding to the determined neural network-based encoder, wherein the components for communication include: Components for receiving the message from the network device; and A component for demodulating the message using the determined neural network-based decoder.

69. The apparatus according to claim 68, wherein, The neural network-based decoder is determined at least in part based on the neural network-based decoder index indicated by the license.

70. The apparatus according to claim 59, wherein, The licenses include dynamic uplink licenses, configured licenses, dynamic downlink licenses, semi-persistent scheduling configurations, or combinations thereof.

71. An apparatus for wireless communication at a network device, comprising: A component used to send permission to user equipment (UE) indicating size parameters for communication messages; A component for determining the neural network-based encoder corresponding to the size parameter from a plurality of neural network-based encoders configured for the UE; as well as A component for communicating the message with the UE according to the permission and at least in part based on the determined neural network-based encoder.

72. The apparatus according to claim 71, wherein, The communication component includes: Components for communicating the message in an incoherent transmission, the incoherent transmission including the message and without a reference signal associated with the message for channel estimation.

73. The apparatus according to claim 71, wherein, The components used for determining include: Components for determining subsets of the plurality of neural network-based encoders corresponding to the size parameters; and A component for determining the neural network-based encoder from the subset of the plurality of neural network-based encoders, at least in part, based on the neural network-based encoder index indicated by the license.

74. The apparatus according to claim 71, wherein, The components used for determining include: Components for determining subsets of the plurality of neural network-based encoders corresponding to the size parameters; and Components for determining the neural network-based encoder from the subset of the plurality of neural network-based encoders based at least in part on transmission parameters.

75. The apparatus according to claim 74, wherein, The transmission parameters include bandwidth indicator, modulation scheme, channel coding rate, or a combination thereof.

76. The apparatus of claim 71, further comprising: A component for sending a configuration message to the UE indicating the plurality of neural network-based encoders and one or more size parameters corresponding to each of the plurality of neural network-based encoders.

77. The apparatus according to claim 71, wherein, The size parameters correspond to the transport block size for the message, the number of resource elements for sending the message, the number of physical resource blocks for sending the message, the number of orthogonal frequency domain multiplexing symbols for sending the message, the number of coded bits for sending the message, or a combination thereof.

78. The apparatus according to claim 71, wherein, The communication component includes: Components for modulating the message using the determined neural network-based encoder; and A component used to send modulated messages to the UE.

79. The apparatus of claim 71, further comprising: Components for determining a neural network-based decoder corresponding to the determined neural network-based encoder, wherein the components for communication include: Components for receiving the message from the UE; and A component for demodulating the message using the determined neural network-based decoder.

80. The apparatus according to claim 71, wherein, The licenses include dynamic uplink licenses, configured licenses, dynamic downlink licenses, semi-persistent scheduling configurations, or combinations thereof.

81. An apparatus for wireless communication, comprising: A component used to determine multiple segments of a set of resources allocated for communication messages; For each of the plurality of segments, the corresponding neural network-based encoder component is determined, at least in part, based on the corresponding size parameters of each segment; as well as A component for communicating the messages in the resource set based at least in part on a corresponding neural network-based encoder determined for each of the plurality of segments.

82. The apparatus of claim 81, further comprising: The component for determining that the transport block size for the message satisfies a threshold size, wherein the plurality of segments are determined at least in part based on the transport block size for the message satisfying the threshold size.

83. The apparatus according to claim 81, wherein, The communication component includes: Components for modulating the message in each of the plurality of segments using a determined corresponding neural network-based encoder for each of the plurality of segments; and A component used to send modulated messages.

84. The apparatus of claim 81, further comprising: A component for determining, for each of the plurality of segments, a corresponding neural network-based decoder corresponding to the determined corresponding neural network-based encoder for each of the plurality of segments, wherein the component for communication includes: Components for receiving the message; and A component for demodulating the message in each of the plurality of segments using a determined corresponding neural network-based decoder for each of the plurality of segments.

85. The apparatus according to claim 81, wherein, The plurality of segments are determined at least in part based on one or more resource elements of the resource set, one or more physical resource blocks of the resource set, one or more subbands of the resource set, one or more orthogonal frequency domain multiplexing symbols of the resource set, one or more groups of orthogonal frequency domain multiplexing symbols of the resource set, or a combination thereof.

86. The apparatus according to claim 81, wherein, The corresponding size parameters of the segments in the plurality of segments correspond to the number of resource elements of the segment, the number of physical resource blocks of the segment, the number of orthogonal frequency domain multiplexing symbols of the segment, the number of coded bits of the segment, or a combination thereof.

87. The apparatus according to claim 81, wherein, The first neural network-based encoder for the first segment of the plurality of segments is different from the second neural network-based encoder for the second segment of the plurality of segments.

88. A non-transitory computer-readable medium storing code for wireless communication at a user equipment (UE), the code comprising instructions executable by a processor for the following operations: Receive permission from network devices to indicate size parameters used for communication messages; Determine the neural network-based encoder corresponding to the size parameter from among a plurality of neural network-based encoders configured at the UE; and The message is communicated with the network device in accordance with the permission and at least in part based on the determined neural network-based encoder.

89. A non-transitory computer-readable medium storing code for wireless communication at a network device, the code comprising instructions executable by a processor for the following operations: Send permission to the user equipment (UE) indicating the size parameters used for communication messages; Determine the neural network-based encoder corresponding to the size parameter from among a plurality of neural network-based encoders configured for the UE; and The message is communicated to the UE in accordance with the permission and at least in part based on the determined neural network-based encoder.

90. A non-transitory computer-readable medium storing code for wireless communication, said code comprising instructions executable by a processor to perform the following operations: Identify multiple segments of the resource set allocated for communication messages; For each of the plurality of segments, a corresponding neural network-based encoder is determined, at least in part, based on the corresponding size parameters of each segment; and The messages are communicated in the resource set based at least in part on a corresponding neural network-based encoder determined for each of the plurality of segments.

Images