Techniques for indicating signal processing procedures for network deployed neural network models

Abstract

Methods, systems, and devices for wireless communications are described. In some examples, a device (e.g., a user equipment (UE)) may obtain a configuration message indicating one or more neural network models from a base station. The UE may then obtain an indication of a sequence of operations for a signal processing procedure for a neural network model of the one or more neural network models. In some examples, the signal processing procedure includes an input pre¬processing procedure or an output pre-processing procedure. Upon obtaining a signal from a base station, the UE may perform the signal processing procedure on the received signal for the neural network model according to the sequence of operations.

Description

[Technical Field]

[0001] This patent application claims the benefit of Indian Provisional Patent Application No. 202121018558, filed on April 22, 2021, entitled “TECHNIQUES FOR INDICATING SIGNAL PROCESSING PROCEDURES FOR NETWORK DEPLOYED NEURAL NETWORK MODELS”, which has been assigned to the assignee of this application.

[0002] The following text relates to wireless communications, and more specifically, to methods and systems for indicating information related to signal processing procedures used in neural network models. [Previous Technology]

[0003] Wireless communication systems are widely deployed to provide various types of communication content such as voice, video, packet data, messaging, and broadcasting. These systems can support communication with multiple users by sharing available system resources (e.g., time, frequency, and power). Examples of such multiplexing access systems include fourth-generation (4G) systems (e.g., Long Term Evolution (LTE) systems, Improved LTE (LTE-A) systems, or LTE-A Pro systems) and fifth-generation (5G) systems (which may be referred to as New Radio (NR) systems). These systems can employ technologies such as Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), Orthogonal FDMA (OFDMA), or Discrete Fourier Transform Spread Orthogonal Frequency Division Multiplexing (DFT-S-OFDM). A wireless multiplexing access communication system may include one or more base stations or one or more network access nodes, each base station or network access node simultaneously supporting communication with multiple communication devices (which may also be referred to as user equipment (UE)). [Summary of the Invention]

[0004] A method for wireless communication at a device in a wireless network is described. The method may include the steps of: obtaining configuration information for the device, wherein the configuration information indicates one or more neural network models for the device; and obtaining an instruction for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model. The method may also include the step of: executing the signal processing procedure for the at least one neural network model using signals obtained at the device, according to the operation sequence.

[0005] An apparatus for wireless communication at a device in a wireless network is described. The apparatus may include a processor and memory coupled to the processor. The processor may be configured to: obtain configuration information for the device, wherein the configuration information indicates one or more neural network models for the device; and obtain an instruction for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model. The processor may also be configured to: execute the signal processing procedure for the at least one neural network model using signals obtained at the device, according to the operation sequence.

[0006] Another apparatus for wireless communication at a device in a wireless network is described. The apparatus may include: means for obtaining configuration information for the device, wherein the configuration information indicates one or more neural network models for the device; and means for obtaining an instruction for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model. The apparatus may also include: means for performing the signal processing procedure for the at least one neural network model using signals obtained at the device, according to the operation sequence.

[0007] A non-transitory computer-readable medium is described, storing code for wireless communication at a device in a wireless network. The code may include instructions executable by a processor to: obtain configuration information for the device, wherein the configuration information indicates one or more neural network models for the device; and obtain instructions for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model. The code may also include instructions executable by the processor to: perform the signal processing procedure for the at least one neural network model using signals obtained at the device, according to the operation sequence.

[0008] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: obtaining a signal transmission configuring the device with a set of operations, the set of operations including one or more operations for the sequence of operations for the at least one neural network model.

[0009] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, obtaining the indication of the sequence of operations may include operations, features, components or instructions for performing the following: obtaining the indication of the sequence of operations for the signal processing program in the configuration message, wherein the configuration message includes a set of operations that includes all operations of the sequence of operations for the at least one neural network model.

[0010] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, obtaining the instruction on the sequence of operations may include operations, features, components or instructions for performing the following: obtaining a second configuration message for the device, wherein the second configuration message indicates: the at least one neural network model, the instruction on the sequence of operations, and a set of operations including all operations for the at least one neural network model.

[0011] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, obtaining the instruction for the sequence of operations may include operations, features, components or instructions for performing: obtaining a set of input parameters, a set of output parameters, or both of the operations for the at least one neural network model.

[0012] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: obtaining an indication of a mapping between one or more neural network models and a set of operating conditions, wherein the execution of the signal processing procedure for the at least one neural network model may be based on the mapping between the one or more neural network models and the set of operating conditions, according to the sequence of operations, using the signal obtained at the device.

[0013] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the set of operating conditions includes a signal-to-noise ratio (SNR) range, bandwidth range, signal scaling range, channel delay profile, signal peak range, or any combination thereof associated with the signal obtained at the device.

[0014] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: outputting a message indicating that the device supports the capability of one or more operations for one or more signal processing programs, wherein the indication of the sequence of operations for the signal processing program for the at least one neural network model may be obtained based on the capability of the device.

[0015] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the sequence of operations includes one or more operations supported by the device.

[0016] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the message indicating that the device supports the capability for one or more operations of the one or more signal processing programs includes: an indication of a threshold input dimension for each of the one or more operations of the one or more signal processing programs or a threshold runtime for each of the one or more operations of the one or more signal processing programs.

[0017] In some instances of the methods, apparatus and nontransitory computer-readable media described herein, obtaining the indication to the sequence of operations may include operations, features, components or instructions for obtaining a Radio Resource Control (RRC) signaling or Media Access Control (MAC) control element (MAC-CE) that includes the indication to the sequence of operations.

[0018] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: obtaining instructions on one or more data formats associated with one or more operations in the sequence of operations, wherein the one or more data formats include Extensible Markup Language (XML) data format, JavaScript Object Markup (JSON) data format, or any combination thereof.

[0019] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the apparatus includes a UE, a base station, a network entity, a relay device, a sidelink device, or an integrated access and backhaul (IAB) node.

[0020] A method for wireless communication at a network entity is described. The method may include the steps of: outputting a configuration message to a device, wherein the configuration message indicates one or more neural network models for the device; and outputting an instruction for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model. The method may also include the step of: outputting a signal to the device based on the instruction for the operation sequence of the signal processing procedure for the at least one neural network model.

[0021] An apparatus for wireless communication at a network entity is described. The apparatus may include a processor and memory coupled to the processor. The processor may be configured to: output a configuration message to a device, wherein the configuration message indicates one or more neural network models for the device; and output an instruction for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model. The processor may also be configured to: output a signal to the device based on the instruction for the operation sequence of the signal processing procedure for the at least one neural network model.

[0022] Another apparatus for wireless communication at a network entity is described. The apparatus may include: means for outputting configuration messages to a device, wherein the configuration messages indicate one or more neural network models for the device; and means for outputting an instruction on an operation sequence of a signal processing procedure for at least one of the one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model. The apparatus may also include: means for outputting a signal to the device based on the instruction on the operation sequence of the signal processing procedure for the at least one neural network model.

[0023] A non-transitory computer-readable medium is described, storing code for wireless communication at a network entity. The code may include instructions executable by a processor to: output a configuration message to a device, wherein the configuration message indicates one or more neural network models for the device; and output instructions for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model. The code may also include instructions executable by the processor to: output a signal to the device based on the instruction for the operation sequence of the signal processing procedure for the at least one neural network model.

[0024] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: outputting a signal transmission configuring the device with a set of operations, the set of operations including one or more operations for the sequence of operations for the at least one neural network model.

[0025] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: obtaining a second sequence of operations for executing a second signaling procedure at the network entity.

[0026] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the output of the instruction to the sequence of operations may include operations, features, components or instructions for performing the following: outputting the instruction to the sequence of operations for the signal processing program in the configuration message, wherein the configuration message includes a set of operations that includes all operations of the sequence of operations for the at least one neural network model.

[0027] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the output of the instruction to the sequence of operations may include operations, features, components or instructions for performing the following: outputting a second configuration message to the device, wherein the second configuration message indicates: the at least one neural network model, the instruction to the sequence of operations, and a set of operations including all operations for the sequence of operations for the at least one neural network model.

[0028] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the instruction to output the sequence of operations may include operations, features, components or instructions for performing one or more operations of the sequence of operations for the at least one neural network model, a set of input parameters, a set of output parameters, or both.

[0029] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: outputting an indication of the mapping between the one or more neural network models and the set of operating conditions.

[0030] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the set of operating conditions includes an SNR range, bandwidth range, signal scaling range, channel delay profile, signal peak range, or any combination thereof associated with the signal transmitted to the device.

[0031] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: obtaining a message indicating that the device supports one or more operations for one or more signal processing programs, wherein obtaining the instruction for the sequence of operations of the signal processing program for the at least one neural network model may be based on the capability of the device.

[0032] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the sequence of operations includes one or more operations supported by the device.

[0033] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the message indicating that the device supports the capability for one or more operations of the one or more signal processing programs includes: an indication of a threshold input dimension for each of the one or more operations of the one or more signal processing programs or a threshold runtime for each of the one or more operations of the one or more signal processing programs.

[0034] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the indication of the operation sequence output may include an operation, feature, component or instruction for performing an RRC signal transmission or MAC-CE that outputs the indication of the operation sequence.

[0035] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: outputting instructions for one or more data formats associated with one or more operations in the sequence of operations, wherein the one or more data formats include XML data formats, JSON data formats, or any combination thereof.

[0036] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the apparatus includes a UE, a base station, a network entity, a relay device, a sidelink device, or an IAB node.

[0037] A method for wireless communication at a device in a wireless network is described. The method may include the steps of: receiving a configuration message for the device, the configuration message indicating one or more neural network models for the device; receiving an instruction for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and executing the signal processing procedure for the at least one neural network model using signals received at the device, according to the operation sequence.

[0038] An apparatus for wireless communication at a device in a wireless network is described. The apparatus may include a processor, memory coupled to the processor, and instructions stored in the memory. The instructions may be executable by the processor to cause the apparatus to perform the following operations: receiving configuration messages for the device, the configuration messages indicating one or more neural network models for the device; receiving instructions for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and, according to the operation sequence, performing the signal processing procedure for the at least one neural network model using signals received at the device.

[0039] Another apparatus for wireless communication at a device in a wireless network is described. The apparatus may include: means for receiving configuration messages for the device, the configuration messages indicating one or more neural network models for the device; means for receiving an instruction for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and means for executing the signal processing procedure for the at least one neural network model using signals received at the device, according to the operation sequence.

[0040] A non-transitory computer-readable medium is described, storing code for wireless communication at a device in a wireless network. The code may include instructions executable by a processor to: receive a configuration message for the device, the configuration message indicating one or more neural network models for the device; receive instructions for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and, according to the operation sequence, execute the signal processing procedure for the at least one neural network model using signals received at the device.

[0041] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: receiving a signal transmission configuring the device with a set of operations, the set of operations including one or more operations for the sequence of operations for the at least one neural network model.

[0042] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: receiving an indication in a configuration message for the sequence of operations for the signal processing procedure, wherein the configuration message includes a set of operations that includes all operations of the sequence of operations for the at least one neural network model.

[0043] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: receiving a second configuration message for the device, the second configuration message indicating: the at least one neural network model, the indication of the sequence of operations, and a set of operations including all operations for the at least one neural network model.

[0044] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: receiving a set of input parameters, a set of output parameters, or both of the operation sequence for the at least one neural network model.

[0045] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: receiving an instruction for a mapping between one or more neural network models and a set of operating conditions, wherein, according to the sequence of operations, performing the signal processing procedure for the at least one neural network model using the signal received at the device may be based on the mapping between the one or more neural network models and the set of operating conditions.

[0046] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the set of operating conditions includes an SNR range, bandwidth range, signal scaling range, channel delay profile, signal peak range, or any combination thereof associated with the signal received at the device.

[0047] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: transmitting a message indicating that the device supports the capability of one or more operations for one or more signal processing programs, wherein receiving the instruction for the sequence of operations of the signal processing program for the at least one neural network model may be based on the capability of the device.

[0048] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the sequence of operations includes one or more operations supported by the device.

[0049] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the message indicating that the device supports the capability for one or more operations of one or more signal processing programs includes: an indication of a threshold input dimension for each of the one or more operations of the one or more signal processing programs or a threshold runtime for each of the one or more operations of the one or more signal processing programs.

[0050] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: receiving an RRC signal transmission or MAC-CE including the indication of the sequence of operations.

[0051] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: receiving instructions for one or more data formats associated with one or more operations in the sequence of operations, including XML data formats, JSON data formats, or any combination thereof.

[0052] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the apparatus includes a UE, a base station, a network entity, a relay device, a sidelink device, or an IAB node.

[0053] A method for wireless communication at a base station is described. The method may include the steps of: transmitting a configuration message to a device, the configuration message indicating one or more neural network models for the device; transmitting an instruction for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and transmitting a signal to the device based on the instruction for the operation sequence of the signal processing procedure for the at least one neural network model.

[0054] An apparatus for wireless communication at a base station is described. The apparatus may include a processor, memory coupled to the processor, and instructions stored in the memory. These instructions may be executable by the processor to cause the apparatus to: transmit configuration information to a device, the configuration information indicating one or more neural network models for the device; transmit an instruction for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and transmit a signal to the device based on the instruction for the operation sequence of the signal processing procedure for the at least one neural network model.

[0055] Another apparatus for wireless communication at a base station is described. The apparatus may include: means for transmitting configuration information to a device, the configuration information indicating one or more neural network models for the device; means for transmitting an instruction on an operation sequence of a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and means for transmitting a signal to the device based on the instruction on the operation sequence of the signal processing procedure for the at least one neural network model.

[0056] 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: transmit a configuration message to a device indicating one or more neural network models for the device; transmit instructions for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and transmit a signal to the device based on the instruction for the operation sequence of the signal processing procedure for the at least one neural network model.

[0057] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: transmitting a signal transmission that configures the device with a set of operations, the set of operations including one or more operations for the sequence of operations for the at least one neural network model.

[0058] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for receiving a second sequence of operations for execution of a second signaling procedure at the base station.

[0059] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: transmitting in the configuration message an indication of the sequence of operations for the signal processing program, wherein the configuration message includes a set of operations that includes all operations of the sequence of operations for the at least one neural network model.

[0060] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: transmitting a second configuration message to the device, the second configuration message indicating the at least one neural network model, the indication of the operation sequence, and a set of operations including all operations for the at least one neural network model.

[0061] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: transmitting a set of input parameters, a set of output parameters, or both of the operation sequence for the at least one neural network model.

[0062] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for transmitting instructions on the mapping between one or more neural network models and sets of operating conditions.

[0063] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the set of operating conditions includes the SNR range, bandwidth range, signal scaling range, channel delay profile, signal peak range, or any combination thereof associated with the signal transmitted to the device.

[0064] Some examples of the methods, apparatuses and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: receiving a message indicating that the device supports the capability of one or more operations for one or more signal processing programs, wherein receiving the instruction for the sequence of operations of the signal processing program for the at least one neural network model may be based on the capability of the device.

[0065] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the sequence of operations includes one or more operations supported by the device.

[0066] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the message indicating that the device supports the capability for the one or more operations of the one or more signal processing programs includes: an indication of a threshold input dimension for each of the one or more operations of the one or more signal processing programs or a threshold runtime for each of the one or more operations of the one or more signal processing programs.

[0067] Some examples of the methods, apparatus and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: transmission of RRC signaling or MAC-CE including the indication of the sequence of operations.

[0068] Some examples of the methods, apparatus and nontransitory computer-readable media described herein may also include operations, features, components or instructions for performing the following: transmitting instructions on one or more data formats associated with one or more operations in the sequence of operations, including XML data formats, JSON data formats, or any combination thereof.

[0069] In some instances of the methods, apparatuses and nontransitory computer-readable media described herein, the apparatus includes a UE, a base station, a network entity, a relay device, a sidelink device, or an IAB node.

Implementation Method

[0081] Some wireless communication systems can support machine learning or neural network models (also known as machine learning models), which can be used to optimize wireless communication processes, such as decoding, encoding, analog-to-digital conversion, generating information to report to higher layers or respond to received signals for transmission, etc. To utilize neural network models, a device (e.g., a UE) can perform preprocessing on received signals before inputting them into the neural network model, which can convert the received signals into a format compatible with the neural network model (e.g., a format that can be received and processed by the neural network model). As used herein, the term "preprocessing" can be used to represent any operation, procedure, algorithm, or mathematical calculation that can be performed to convert a signal into a format that can be received and processed by the neural network model.

[0082] Similarly, the device can perform post-processing on the output of the neural network model, which can convert the output into a format compatible with the output reported to higher layers of the network or device (e.g., a format of output that can be received and / or processed by the network or higher layers). The output of the neural network model may include modified versions of signals input to the neural network model, decisions or calculations performed by the neural network model based on the input, etc. However, methods for providing information and instructions to wireless devices (e.g., UEs) for performing signal processing (e.g., preprocessing and postprocessing) for a neural network model implemented within a wireless communication system have not yet been considered. That is, some wireless communication systems have not defined or envisioned signal transmission and configuration that can be used to provide information (e.g., operation, instructions) to UEs and other wireless devices, such signal transmission and configuration enabling the respective UEs and wireless devices to execute neural network models within the respective wireless communication systems.

[0083] As described herein, a device may receive (e.g., acquire) signal transmissions instructing signal processing (e.g., preprocessing or postprocessing) operations, and commands for performing operations for corresponding preprocessing, postprocessing, and neural network models. Neural network models may include the same signal processing operations (e.g., basic functions or non-trainable layers), such as signal scaling operations, cyclic shift operations, inverse fast Fourier transform (IFFT) operations, etc. Although signal processing operations may be common across neural network models, the order in which operations are performed and the input and output parameters used for the operations may differ for each neural network model. Therefore, various aspects of this application relate to signal transmissions and techniques that enable a wireless device (e.g., a UE) to be configured with preprocessing and postprocessing operations associated with signal processing and neural network models. In this application, various aspects of this application enable the UE to perform operations associated with signal processing and neural network models in the correct order (e.g., the correct temporal order of operations), thereby enabling the wireless device to perform signal processing operations associated with neural network models more efficiently and effectively.

[0084] For the purposes of this case, the terms "signal processing procedure" and "signal processing operation" may be used to refer to any procedure or operation for processing physical layer signals covering the time domain, frequency domain, spatial domain, and / or code domain as observed at one or more time instances. Therefore, the terms "signal processing procedure" and "signal processing operation" may be used to process various types of signals, including radio frequency signals, audio / video (A / V) signals, time series data, images, etc.

[0085] In one instance, a network (e.g., an entity of a wireless communication system, such as a network entity or base station) may configure a device with a set of signal processing operations (e.g., which may be defined in a standard) and a sequence of signals indicating the operations for each neural network model (e.g., commands for executing the set of operations) and input and output parameters for the operations (e.g., via RRC signaling or MAC-CE). The operation sequence may indicate a sequence of subsets of operations included in the set of operations (e.g., the set of operations includes a first subset of operations, followed by a second subset of operations, etc.) and / or a sequence of operations within a given subset of operations (e.g., the subset of operations includes a first operation, followed by a second operation). After receiving or acquiring a signal (e.g., from a network entity), the device may perform signal processing according to the operation sequence. In another instance, a network (e.g., a network entity) may signal the signal processing operations, operation sequences, input and output parameters, and the neural network model as part of a complete packet to the device.

[0086] In some instances, the device may also acquire or receive signaling transmissions indicating a relationship between the neural network model and its operating range (e.g., SNR range or bandwidth range), which the device can use to dynamically select a neural network model for application to received signals. Furthermore, the device may output or transmit signaling transmissions to the network indicating its ability to perform signal processing operations, which the network can use to determine which neural network model to provide to the device. Using the methods described herein, the device can acquire (e.g., receive) information related to signal processing and perform signal processing based on information for a neural network model deployed by a wireless device (e.g., a UE) within a wireless communication system. Therefore, the techniques described herein enable UEs and other wireless devices to perform signal processing operations that facilitate neural network models more efficiently and effectively, enabling more complex and reliable processing within a wireless communication system.

[0087] First, various forms of the present invention are described in the context of a wireless communication system. Additional forms of the present invention are described in the context of flowcharts, machine learning processes, and process flows. Various forms of the present invention are further illustrated by apparatus diagrams, system diagrams, and flowcharts relating to techniques for instructing signal processing procedures for deploying neural network models on a network, and are described with reference to these diagrams.

[0088] Figure 1 illustrates an example of a wireless communication system 100, according to one or more embodiments of the present invention, supporting techniques for instructing signal processing procedures for deploying neural network models on a network. The wireless communication system 100 may include one or more network entities 105, one or more UEs 115, and a core network 130. In some instances, the wireless communication system 100 may be a Long Term Evolution (LTE) network, an improved LTE (LTE-A) network, an LTE-A Pro network, or a New Radio (NR) network. In some instances, the wireless communication system 100 may support enhanced broadband communication, ultra-reliable (e.g., mission-critical) communication, low-latency communication, or communication with low-cost and low-complexity devices, or any combination thereof.

[0089] Network entity 105 can be distributed throughout a geographical area to form wireless communication system 100, and can be devices of different forms or with different capabilities. Network entity 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 in which base station 105 and UE 115 can support signal transmission according to one or more radio access technologies.

[0090] UE 115 can be distributed throughout the entire coverage area 110 of the wireless communication system 100, and each UE 115 can be stationary, active, or both at different times. UE 115 can be devices of different forms or with different capabilities. Some exemplary UE 115s are illustrated in Figure 1. The UE 115 described herein is capable of communicating with various types of devices, such as other UE 115s, network entities 105, or network devices (e.g., core network nodes, relay devices, IAB nodes, or other network devices), as shown in Figure 1.

[0091] Network entity 105 may communicate with core network 130, or communicate with each other, or perform both of the above operations. For example, network entity 105 may interface with core network 130 via one or more backhaul links 120 (e.g., via S1, N2, N3, or other interfaces). Network entity 105 may communicate with each other directly (e.g., directly between network entities 105) on backhaul links 120 (e.g., via X2, Xn, or other interfaces), or indirectly (e.g., via core network 130), or perform both of the above operations. In some instances, backhaul link 120 may be one or more radio links or may include one or more radio links.

[0092] One or more base stations 105 in the network entity 105 described herein may include or may be referred to by those of ordinary skill as base station transceiver, radio base station, access point, radio transceiver, node B, evolved node B (eNB), next-generation node B or gigabit node B (any of which may be referred to as gNB), home node B, home evolved node B, or some other suitable term.

[0093] As described herein, a node (which may be referred to as a node, network node, network entity, or wireless node) may be a base station (e.g., any base station described herein), a UE (e.g., any UE described herein), a network controller, apparatus, device, computing system, one or more elements, and / or another suitable processing entity configured to perform any of the techniques described herein. For example, a network node may be a UE. As another example, a network node may be a base station. As another example, a first network node may be configured to communicate with a second or third network node. In one embodiment of this example, the first network node may be a UE, the second network node may be a base station, and the third network node may be a UE. In another embodiment of this example, the first network node may be a UE, the second network node may be a base station, and the third network node may be a base station. In other embodiments of this example, the first, second, and third network nodes may differ from those embodiments. Similarly, references to UE, base station, device, equipment, computing system, etc., may include disclosing UE, base station, device, equipment, computing system, etc., as network nodes. For example, a disclosure that a UE is configured to obtain or receive information from a base station also discloses that a first network node is configured to receive information from a second network node. Consistent with the content of this application, once a specific instance is broadened according to the content of this application (e.g., a UE being configured to receive information from a base station also discloses that a first network node is configured to receive information from a second network node), a broader instance of a narrower instance can be interpreted in reverse, but in a broad, open-ended manner. In the above-described instance where a UE is configured to receive information from a base station also discloses that a first network node is configured to receive information from a second network node, the first network node may refer to a first UE, a first base station, a first device, a first equipment, a first computing system, a first one or more components, a first processing entity, etc., configured to receive information; and the second network node may represent a second UE, a base station, a second device, a second equipment, a second computing system, a first one or more components, a first processing entity, etc.

[0094] As described herein, different terms may be used to describe the communication of information (e.g., any information, signal, etc.) in various forms. The disclosure of one communication term includes the disclosure of other communication terms. For example, a first network node may be described as being configured to output or transmit information to a second network node. In this example and consistent with the present invention, the disclosure of a first network node being configured to transmit information to a second network node includes the disclosure of a first network node being configured to provide, send, output, transmit, or transfer information to a second network node. Similarly, in this example and consistent with the present invention, the disclosure of a first network node being configured to transmit information to a second network node includes the disclosure of a second network node being configured to receive, obtain, or decode information provided, sent, output, transmitted, or transferred by the first network node.

[0095] UE 115 may include or be referred to as a mobile device, wireless device, remote device, handheld device, or user equipment, or some other suitable term, wherein "device" may also be referred to as a cell, station, terminal, or client, and other instances. 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 instances, among others, UE 115 may include or be referred to as a wireless area loop (WLL) station, Internet of Things (IoT) device, Internet of Everything (IoE) device, or machine-type communication (MTC) device, which, among other instances, may be implemented in various objects such as electrical appliances, vehicles, or instruments.

[0096] The UE 115 described herein is capable of communicating with various types of devices, such as other UEs 115 that may sometimes act as repeaters, as well as network entities 105 and network devices, including, among other instances, macro eNBs or gNBs, small cell eNBs or gNBs, or relay base stations, as shown in Figure 1. In some instances, the UE 115 may communicate with the core network 130 via communication link 155.

[0097] UE 115 and network entity 105 can communicate wirelessly with each other via one or more communication links 125 on one or more carriers. The term "carrier" represents a set of radio frequency spectrum resources having a defined entity layer structure for supporting communication link 125. For example, a carrier for communication link 125 may include a portion of a radio frequency spectrum band (e.g., a bandwidth portion) that operates according to one or more entity layer channels for a given radio access technology (e.g., LTE, LTE-A, LTE-A Pro, NR). Each entity layer channel may carry acquisition signaling (e.g., synchronization signals, system information), control signaling for coordinating the operation of 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 may 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) component carriers and time division duplex (TDD) component carriers.

[0098] As described herein, different terms may be used to describe the communication of information (e.g., any information, signal, etc.) in various forms. The disclosure of one communication term includes the disclosure of other communication terms. For example, a first network node may be described as being configured to transmit information to a second network node. In this example and consistent with the present invention, the disclosure of a first network node being configured to transmit information to a second network node includes the disclosure of a first network node being configured to provide, send, output, transmit, or transfer information to a second network node. Similarly, in this example and consistent with the present invention, the disclosure of a first network node being configured to transmit information to a second network node includes the disclosure of a second network node being configured to receive, obtain, or decode information provided, sent, output, transmitted, or transferred by the first network node.

[0099] The signal waveform transmitted on the carrier can be composed of multiple subcarriers (e.g., using multicarrier modulation (MCM) techniques such as orthogonal frequency division multiplexing (OFDM) or discrete Fourier transform spread spectrum OFDM (DFT-S-OFDM). In a system employing MCM techniques, 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 the 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 decoding rate of the modulation scheme, or both). Therefore, the more resource elements the UE 115 receives and the higher the order of the modulation scheme, the higher the data rate for the UE 115 can be. Radio communication resources can represent a combination of radio frequency spectrum resources, temporal resources, and spatial resources (e.g., spatial layers or beams), and the use of multiple spatial layers can further increase the data rate or data integrity for communication with the UE 115.

[0100] The time intervals for network entity 105 or UE 115 can be represented as multiples of a basic time unit (which may, for example, refer to a sampling period in seconds, where the maximum supported subcarrier interval can be represented, and the maximum supported discrete Fourier transform (DFT) size can be represented). The time intervals of communication resources can be organized according to radio frames, each having a specified duration (e.g., 10 milliseconds (ms)). Each radio frame can be identified via a system frame number (SFN) (e.g., ranging from 0 to 1023).

[0101] Each frame may include multiple consecutively numbered subframes or time slots, and each subframe or time slot may have the same duration. In some instances, 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., the number depends on the length of the cyclic prefix added before each symbol period). In some wireless communication systems 100, time slots may be further divided into multiple microtime slots containing one or more symbols. Excluding the cyclic prefix, each symbol period may contain one or more (e.g., ) sampling periods. The duration of a symbol period may depend on the subcarrier spacing or the operating frequency band.

[0102] A subframe, time slot, micro-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 instances, the duration of the TTI (e.g., the number of symbol periods in the TTI) can be variable. Alternatively, the smallest scheduling unit of the wireless communication system 100 can be dynamically selected (e.g., in a short pulse of a shortened TTI (sTTI)).

[0103] Entity 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 entity control channels and entity data channels on a downlink carrier. A control region (e.g., a control resource set (CORESET)) for an entity control channel can be defined by multiple symbol periods and can extend over the system bandwidth or a subset of the system bandwidth of the carrier. One or more control regions (e.g., CORESETs) can be configured for a set of UEs 115. For example, one or more UEs in UE 115 can monitor or search for control regions 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 for control channel candidates can represent the number of control channel resources (e.g., control channel elements (CCEs)) associated with coded information for a control information format having a given payload size. The search space set may include a shared search space set configured to send control information to multiple UEs 115 and a UE-specific search space set configured to send control information to a particular UE 115.

[0104] In some instances, base station 105 may be mobile, and therefore provide communication coverage for mobile geographic coverage areas 110. In some instances, different geographic coverage areas 110 associated with different technologies may overlap, but the different geographic coverage areas 110 may be supported by the same base station 105. In other instances, overlapping geographic coverage areas 110 associated with different technologies may be supported by different network entities 105. Wireless communication system 100 may include, for example, heterogeneous networks, where different types of network entities 105 use the same or different radio access technologies to provide coverage for various geographic coverage areas 110.

[0105] Some UEs 115 (e.g., MTC or IoT devices) can be low-cost or low-complexity devices and can provide automated machine-to-machine communication (e.g., via machine-to-machine (M2M) communication). M2M communication or MTC can represent data communication technologies that allow devices to communicate with each other or with base station 105 without human intervention. In some instances, M2M communication or MTC can include communication from a device that integrates sensors or meters to measure or capture information and relays such information to a central server or application that uses the information or presents it to a human interacting with the application. Some UEs 115 can be designed to collect information or automate the behavior of machines or other devices. Examples of applications for MTC devices include smart metering, inventory monitoring, water level monitoring, equipment monitoring, healthcare monitoring, wildlife monitoring, climate and geological event monitoring, fleet management and tracking, remote security sensing, physical access control, and transaction-based billing of transfers.

[0106] The wireless communication system 100 may be configured to support ultra-reliable communication or low-latency communication, or various combinations thereof. For example, the wireless communication system 100 may be configured to support ultra-reliable low-latency communication (URLLC) or mission-critical communication. The UE 115 may be designed to support ultra-reliable, low-latency, or mission-critical functions (e.g., mission-critical functions). Ultra-reliable communication may include private or group communication and may 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 may include service prioritization, and mission-critical services may 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.

[0107] In some instances, UE 115 may also communicate directly with other UE 115 on a 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 instances, multiple groups of UE 115s communicating via D2D communication may utilize a one-to-many (1:M) system, where each UE 115 transmits to each other UE 115 in the group. In some instances, base station 105 facilitates the scheduling of resources for D2D communication. In other cases, D2D communication is performed between UE 115s without involving base station 105.

[0108] Core network 130 can provide user authentication, access authorization, tracking, Internet Protocol (IP) connectivity, and other access, routing, or mobility functions. Core network 130 may be an Evolved Packet Core (EPC) or a 5G Core (5GC), which may include at least one control plane entity (e.g., a Mobility Management Entity (MME), an Access and Mobility Management Function Unit (AMF)) for managing access and mobility, and at least one user plane entity (e.g., a Serving Gateway (S-GW), a Packet Data Network (PDN) Gateway (P-GW), or a User Plane Function Unit (UPF)) for routing or interconnecting packets to external networks. The control plane entity may manage non-access stratum (NAS) functions, such as mobility, authentication, and bearer management for UE 115 served by network entity 105 associated with core network 130. User IP packets may be transmitted via the user plane entity, which may provide IP address allocation and other functions. The user plane entity can connect to IP services 150 for one or more network service providers. IP services 150 may include access to the Internet, intranet, IP Multimedia Subsystem (IMS), or packet-switched streaming services.

[0109] Some network devices (e.g., base station 105) may include sub-elements such as access network entity 140, which may be an instance of an access node controller (ANC). Each access network entity 140 may communicate with UE 115 via one or more other access network transport entities 145 (which may be referred to as a radio headend, smart radio headend, 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 headends and ANCs) or incorporated into a single network device (e.g., base station 105).

[0110] The wireless communication system 100 can operate using one or more frequency bands (e.g., in the range of 300 MHz to 300 GHz). In some instances, 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 ranges from approximately one decimeter to one meter in length. UHF waves may be blocked or redirected by buildings and environmental features, but the waves may be sufficient to penetrate structures for macrocells to provide service to the UE 115 located indoors. Compared to the transmission of smaller frequencies and longer waves using the lower frequencies (HF) or ultra-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).

[0111] The electromagnetic spectrum is typically subdivided into various categories, bands, channels, etc., based on frequency / wavelength. In 5G NR, two initial operating bands have been identified as frequency range names FR1 (410 MHz – 7.125 GHz) and FR2 (24.25 GHz – 52.6 GHz). It should be understood that although a portion of FR1 is greater than 6 GHz, in various documents and articles, FR1 is usually (interchangeably) referred to as the "below 6 GHz" band. Similar naming issues sometimes arise with FR2; although FR2 is different from the extremely high frequency (EHF) band (30 GHz – 300 GHz), it is usually (interchangeably) referred to as the "millimeter wave" band in documents and articles, while the EHF band is identified as the "millimeter wave" band by the International Telecommunication Union (ITU).

[0112] The frequencies between FR1 and FR2 are generally referred to as intermediate frequencies (IFs). Recent 5G NR research has identified the operating bands of these IFs as the frequency range name FR3 (7.125 GHz – 24.25 GHz). Bands falling within FR3 can inherit the characteristics of FR1 or FR2, and thus can effectively extend the characteristics of FR1 or FR2 to the IF. Additionally, higher frequency bands are currently being explored to extend 5G NR operation above 52.6 GHz. For example, three higher operating frequency bands have been identified as the frequency range names FR4a or FR4-1 (52.6 GHz – 71 GHz), FR4 (52.6 GHz – 114.25 GHz), and FR5 (114.25 GHz – 300 GHz). Each of these higher frequency bands falls within the EHF band.

[0113] Considering the above, unless otherwise specifically stated, it should be understood that when the term "below 6 GHz" is used herein, it can broadly refer to a frequency that is less than 6 GHz, can be within FR1, or can include an intermediate frequency. Furthermore, unless otherwise specifically stated, it should be understood that when the term "millimeter wave" is used herein, it can broadly refer to a frequency that can include an intermediate frequency, can be within FR2, FR4, FR4-a, or FR4-1 or FR5, or can be within the EHF band.

[0114] The wireless communication system 100 may utilize both licensed and unlicensed radio frequency spectrum bands. For example, the wireless communication system 100 may employ Licensed Assisted Access (LAA), LTE Unlicensed (LTE-U) radio access technology, or NR technology in unlicensed frequency bands (such as the 5 GHz Industrial, Scientific, and Medical (ISM) band). When operating in unlicensed radio frequency spectrum bands, devices (such as network entity 105 and UE 115) may employ carrier sensing for collision detection and avoidance. In some instances, operation in unlicensed frequency bands may be based on carrier aggregation configurations that combine component carriers operating in licensed frequency bands (e.g., LAA). Among other instances, operation in unlicensed spectrum may include downlink transmissions, uplink transmissions, P2P transmissions, or D2D transmissions.

[0115] Base station 105 or UE 115 may be equipped with multiple antennas, which can be used to employ technologies 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 (which may support MIMO operation or transmit or receive beamforming). For example, one or more base station antennas or antenna arrays may be co-located at antenna elements, such as antenna towers. In some instances, 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 rows and columns of antenna ports that base station 105 can use to support beamforming for communication with UE 115. Similarly, UE 115 may have one or more antenna arrays that can support various MIMO or beamforming operations. Alternatively or additionally, antenna panels may support radio frequency beamforming for signals transmitted via antenna ports.

[0116] Beamforming (which may also be referred to 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 or UE 115) to form or guide an antenna beam (e.g., a transmission beam, a receiving beam) along a spatial path between the transmitting and receiving devices. Beamforming can be achieved by combining signals transmitted via antenna elements of an antenna array such that some signals propagating in a particular orientation relative to the antenna array experience constructive interference, while other signals experience destructive interference. Adjustments to the signals transmitted via the antenna elements can include applying amplitude offset, phase offset, or both to the signals carried by the transmitting or receiving device via the antenna elements associated with that device. The adjustments associated with each antenna element can be defined by a set of beamforming weights associated with a particular orientation (e.g., relative to the antenna array of the transmitting or receiving device, or relative to some other orientation).

[0117] The wireless communication system 100 may 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 may be IP-based. The radio link control (RLC) layer may perform packet fragmentation and reassembly for transmission over logical channels. The MAC layer may perform prioritization and multiplexing from logical channel to transport channel. The MAC layer may 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 RRC protocol layer may provide the establishment, configuration, and maintenance of RRC connections between the UE 115 and the base station 105 or core network 130 (which supports radio bearers for user plane data). At the physical layer, transport channels may be mapped to physical channels.

[0118] The techniques described herein, in addition to being performed between UE 115 and network entity 105, or alternatively between UE 115 and network entity 105, can also be implemented via additional or alternative wireless devices (including IAB node 104, distributed unit (DU) 165, centralized unit (CU) 160, radio unit (RU) 170, etc.). For example, in some implementations, the various forms described herein can be implemented in the context of a decomposed radio access network (RAN) architecture (e.g., an open RAN architecture). In the decomposed architecture, the RAN can be separated into three functional areas corresponding to CU 160, DU 165, and RU 170. The functional separation between CU 160, DU 165, and RU 175 is flexible, and therefore results in various permutations of different functions depending on the functions performed at CU 160, DU 165, and RU 175 (e.g., MAC functions, baseband functions, radio frequency functions, and any combination thereof). For example, functional separation of protocol stacking can be used between DU 165 and RU 170, so that DU 165 can support one or more layers of protocol stacking, and RU 170 can support one or more different layers of protocol stacking.

[0119] Some wireless communication systems (e.g., wireless communication system 100), infrastructure, and spectrum resources used for NR access may additionally support wireless backhaul link capabilities to supplement wired backhaul connections, thereby providing an IAB network architecture. One or more network entities 105 may include CU 160, DU 165, and RU 170, and may be referred to as donor network entity 105 or IAB donor. One or more DU 165s (e.g., and / or RU 170) associated with donor base station 105 may be partially controlled by CU 160 associated with donor base station 105. One or more donor network entities 105 (e.g., IAB donors) may communicate with one or more additional network entities 105 (e.g., IAB node 104) via supported access and backhaul links. IAB node 104 may support mobile terminal (MT) functions controlled and / or scheduled by the DU 165 of the coupled IAB donor. Additionally, IAB node 104 may include DU 165, which supports communication links (e.g., downstream) with additional entities (e.g., IAB node 104, UE 115, etc.) within the configuration of the relay chain or access network. In this case, one or more elements of the decomposed RAN architecture (e.g., one or more IAB node 104 or elements of IAB node 104) may be configured to operate according to the techniques described herein.

[0120] In some instances, the wireless communication system 100 may include a core network 130 (e.g., a next-generation core network (NGC)), one or more IAB implementers, IAB nodes 104, and a UE 115, wherein the IAB nodes 104 may be partially controlled by each other and / or the IAB implementers. The IAB implementers and IAB nodes 104 may be various instances of network entity 105. The IAB implementers and one or more IAB nodes 104 may be configured as a relay chain (e.g., or communicate according to a relay chain).

[0121] For example, an access network (AN) or RAN can refer to communication between an access node (e.g., an IAB facility), IAB node 104, and one or more UEs 115. An IAB facility can facilitate connectivity between the core network 130 and the AN (e.g., via a wired or wireless connection to the core network 130). That is, an IAB facility can refer to a RAN node having a wired or wireless connection to the core network 130. An IAB facility may include a CU 160 and at least one DU 165 (e.g., and an RU 170), wherein the CU 160 can communicate with the core network 130 via an NG interface (e.g., a backhaul link). The CU 160 can manage Layer 3 (L3) functions (e.g., RRC, Service Data Adaptation Protocol (SDAP), PDCP, etc.) and signaling. At least one DU 165 and / or RU 170 can manage lower-layer functions and signaling, such as Layer 1 (L1) and Layer 2 (L2) functions (e.g., RLC, MAC, entity (PHY) etc.), and each can be at least partially controlled by CU 160. DU 165 can support one or more different cells. IAB donors and IAB nodes 104 can communicate via an F1 interface according to some protocol used to define signaling messages (e.g., F1 AP protocol). In addition, CU 160 can communicate with the core network via an NG interface (which may be an instance of a backhaul link) and can communicate with other CU 160s (e.g., CU 160 associated with an alternative IAB donor) via an Xn-C interface (which may be an instance of a backhaul link).

[0122] IAB node 104 may refer to a RAN node that provides IAB functions (e.g., access for UE 115, radio self-reload capability, etc.). IAB node 104 may include DU 165 and MT. DU 165 may act as a distributed scheduling node toward child nodes associated with IAB node 104, and MT may act as a scheduled node toward a parent node associated with IAB node 104. That is, an IAB grant may be referred to as a parent node communicating with one or more child nodes (e.g., an IAB grant may relay transmissions for the UE via one or more other IAB nodes 104). In addition, IAB node 104 may also be referred to as a parent node or child node of other IAB nodes 104, depending on the AN's relay chain or configuration. Therefore, the MT entity of IAB node 104 (e.g., MT) can provide a Uu interface for child nodes to receive signal transmissions from parent IAB node 104, and the DU interface (e.g., DU 165) can provide a Uu interface for parent nodes to notify child IAB node 104 or UE 115 with signals.

[0123] For example, IAB node 104 can refer to a parent node associated with the IAB node and a child node associated with the IAB implementer. The IAB implementer may include a CU 160 having a wired (e.g., fiber optic) or wireless connection to the core network and may act as the parent node of IAB node 104. For example, the DU 165 of the IAB implementer may relay communications to UE 115 via IAB node 104 and may directly send transmissions to UE 115 via signaling. The CU 160 of the IAB implementer may signal to IAB node 104 via an F1 interface to notify of the establishment of the communication link, and IAB node 104 may schedule transmissions via DU 165 (e.g., transmissions relayed from the IAB implementer to UE 115). That is, data may be relayed to or from IAB node 104 via signaling through the NR Uu interface of the MT to IAB node 104. Communication with IAB node 104 can be scheduled by IAB implementer DU 165, and communication with IAB node 104 can be scheduled by IAB node 104 DU 165.

[0124] When the techniques described herein are applied to a decomposed RAN architecture, one or more elements of the decomposed RAN architecture (e.g., one or more IAB nodes 104 or elements of IAB node 104) can be configured to support the techniques described herein for large round-trip times in random access channel procedures. For example, some operations described as being performed by UE 115 or base station 105 can be additionally or alternatively performed by elements of the decomposed RAN architecture (e.g., IAB nodes, DUs, CUs, etc.).

[0125] In some instances, the wireless communication system 100 may support neural network modeling, and a communication manager 101 may be included in the device to support signal processing when implementing the neural network model. In some examples, the base station 105 may include a communication manager 101-a, and the UE 115 may include a communication manager 101-b. The communication manager 101-a may transmit one or more neural network models, instructions for operation sequences, and signals to the UE 115. The operation sequences may notify the UE 115 of commands for performing signal transmission processing for each neural network model. In response, the communication manager 101-b may select a neural network model based on the received signals and perform signal transmission processing according to the operation sequence of the selected model. In some instances, the communication manager 101-b may transmit signals configuring the UE 115 with an operation set before transmitting the operation sequence, wherein the operation sequence includes one or more operations from the operation set. In another instance, the communication manager 101-b can transmit an operation set, an operation sequence, and one or more neural network models together as a single packet. In either case, the UE 115 can obtain information related to the signal processing of one or more neural network models.

[0126] Figure 2 illustrates an example of a wireless communication system 200 that supports one or more forms of a technology for instructing signal processing procedures for deploying neural network models in accordance with the present invention. In some instances, the wireless communication system 200 may implement or be implemented by various forms of the wireless communication system 100. For example, the wireless communication system 200 may include a base station 105-a and a UE 115-a, which may be instances of base station 105 and UE 115 as described with reference to Figure 1. In some instances, base station 105-a and UE 115-a may be located in coverage area 110-a and may communicate via downlink communication link 205.

[0127] In some instances, the wireless communication system 200 may support machine learning or neural network models. A neural network model may be an instance of a program trained to recognize patterns, and the wireless communication system may utilize the neural network model to optimize the wireless communication process. For example, the wireless communication system may utilize the neural network model to detect delays related to line-of-sight (LOS) signals, and other instances.

[0128] In some cases, a network (such as base station 105-a) can configure a wireless device (such as UE 115-a) with a neural network model, enabling the wireless device to implement the neural network model. For example, base station 105-a can encode one or more neural network models using a format (e.g., Open Neural Network Switching (ONNX) format) and output (e.g., provide, transmit) them to UE 115-a via downlink communication link 205. UE 115-a can use a decoder to interpret one or more neural network models and implement one or more neural network models. In some instances, the network can determine which neural network model to provide to the wireless device based on the current operating scenario (e.g., number of antennas, operating SNR, operating bandwidth portion, modulation or RF model). Each neural network model provided to the wireless device may be valid within a certain operating range. For example, each neural network model may be valid within a given SNR range, a given bandwidth range, a given channel power delay profile, a given signal scaling range, a given signal peak range, etc. Wireless devices can determine the characteristics of the signals they acquire (e.g., receive) (e.g., signal SNR or signal bandwidth) and implement neural network models that encompass the characteristics of the acquired signals in their operational scope.

[0129] To implement or report the output of a neural network model, the wireless device may perform signal processing. Signal processing may include at least one of preprocessing or postprocessing. The wireless device (e.g., UE 115-a) may perform preprocessing to convert the received signal into a format compatible with the input of the neural network model. Similarly, the wireless device (e.g., UE 115-a) may perform postprocessing to convert the output of the neural network model into a format that can be mapped to a compatible report, wherein the report can be sent to a higher layer (inner or outer layer) of the wireless device or transmitted as a signal to other devices (e.g., base station 105-a).

[0130] Using other techniques, a wireless device (e.g., UE 115-a) can be pre-configured with information about how to perform signal processing on all neural network models designed by the network (which may include hundreds or thousands of neural network models designed for different operating scenarios). However, the network can provide the wireless device with a relatively small subset of such neural network models, and therefore, pre-configuring the wireless device with information about how to perform signal processing for all neural network models may be impractical or inefficient. Furthermore, neural network architectures can continue to evolve, potentially becoming more suitable for different processing operations.

[0131] As described herein, a wireless device may acquire or receive signaling information instructing the execution of signal processing (e.g., preprocessing or postprocessing) for a network-deployed neural network model. In some instances, the operations for signal processing (e.g., basic functions or non-trainable layers) may be common across different neural network models, but the order in which the operations are performed and the input and output parameters may differ. In one instance, UE 115-a or base station 105-a may be pre-configured with a set of operations, where each operation in the set may include several inputs and outputs (options for inputs and outputs).

[0132] Some examples of operations that can be configured for UE 115-a include Channel Feedback Reporting (CFR) operations, zero-padding operations, IFFT operations (e.g., operations used to convert a signal from the frequency domain to the time domain), signal scaling operations (e.g., operations used to scale the size or amplitude of a signal), peak search operations (e.g., operations used to identify the peak or maximum size / amplitude of a signal), cyclic shift operations (e.g., bit-by-bit rotation, or operations used to shift bits of a signal), truncation operations, concatenation operations, complex-to-real operations, any linear algebraic operations (e.g., singular value decomposition (SVD), QR decomposition, Cholesky decomposition, determinant, rank, condition number, or eigenvalue), or any matrix and vector equations. In some instances, matrix and vector operations can be obtained via mathematical or scientific computing libraries (such as NumPy, SciPy, LinAlg, etc.). That is, UE 115-a can utilize any vector, matrix, or tensor operations or any linear algebraic method as part of signal processing (e.g., preprocessing or postprocessing).

[0133] When communicating with base station 105-a, UE 115-a can obtain or receive one or more neural network models from the network. For example, UE 115-a can receive configuration message 210 indicating a first neural network model from base station 105-a. After receiving one or more neural network models, UE 115-a can receive execution sequence indication 215. Execution sequence indication 215 can indicate an operation sequence or command to execute at least a subset of the operation set for one or more signal processing procedures provided to UE 115-a for each neural network model. For example, execution sequence indication 215 for a LOS delay detection neural network model can indicate to execute the operations for signal processing in the following order: channel feedback reporting operation, zero-padding operation, IFFT operation, signal scaling operation, peak search operation, cyclic shift operation, truncation operation, concatenation operation, and realignment operation. Additionally, execution sequence instruction 215 may include instructions for input and output parameters for each step of each neural network model (e.g., for each operation in the sequence). For example, execution sequence instruction 215 may instruct a left peak shift of the input and a peak index for the cyclic shift operation of the LOS delay detection neural network model as the output. In some instances, execution sequence instruction 215 may be included in RRC signaling or MAC-CE. In some instances, signaling formats such as XML and JSON may be used to signal execution sequence instruction 215, or to indicate format support for preprocessing, postprocessing, or machine learning models. Exemplary code for execution sequence instructions for LOS delay detection neural network models is shown in Table 1 below. LOS-Delay-PosNN-Model-PreProc::= SEQUENCE { ParamConfig::= SEQUENCE { PerResourceSet CHOICE{Independent, Combined} PerResource CHOICE{Independent, Combined} PerRxAntennaInput CHOICE{Independent, Combined, Best-N} Best-N CHOICE{1,2,.,NumRxAnt] } ExecuteSequence:: { RemoveZeroFromCFRComb ZeroPad SEQUENCE(2 of INTEGER(1,NumTones)) IFFT SEQUENCE{ { Input-Oversampling INTEGER{1,2,4,8} Output-Scaling CHOICE(1,N,1 / N, 1 / sqrt(N)) Output-DC-Centering CHOICE{True, False} } Scaling SEQUENCE{ Input-ScalingType CHOICE{PeakScaling, L1NormScaling, L2NormScaling} FindPeak SEQUENCE{ Output-PeakIndex INTEGER(1 to LenInput) } CircShift SEQUENCE{ Input-LeftPeakShift Output-PeakIndex } Truncate SEQEUNCE( Input-LeftTruncate INTEGER(1 to LenInput) Input-RightTruncate INTEGER(1 to LenInput) } ComplexToReal SEQUENCE{ Input-ExpandDim INTEGER(2,DimInput+1) Input-AddMagVector CHOICE(True, False) Input-AddPhaseVector CHOICE(True, False) } MultiResourceConcatenate SEQUENCE{ Input-ConcatenateDim INTEGER(1 to DimInput) Input-PRSResourceConcatenate CHOICE(True, False) Input-PRSResourceSetConcatenate CHOICE(True, False) Input-RXAntennaConcatenate CHOICE(True, False)} CheckDim SEQUENCE{ dim1 INTEGER(1 to 65536) dim2 INTEGER(1 to 65536) } NNModel ModelName ShiftAndScaleNNOutput SEQUENCE{ Input-shift MULT(-1, Output-PeakIndex) Input-scale MULT(INTEGER, Input-shift) } QUANTIZE SEQUENCE{ Input-QuantizeType CHOICE{Uniform, Non-Uniform} Input-Stepsize INTEGER(1 to MaxStepSize) Output-Value INTEGER(1 to 65536) } } } Table 1

[0134] The wireless device can perform signal processing on the received signal based on the execution sequence indication 215. For example, UE 115-a can obtain signal 220 from base station 105-a and determine the characteristics of signal 220 (e.g., signal SNR or signal bandwidth). UE 115-a can select a neural network model (e.g., a first neural network model) based on the determined signal characteristics and perform preprocessing on the obtained signal 220 according to the operation sequence indicated in the execution sequence indication 215 for the selected neural network model. Once signal 220 has been preprocessed, UE 115-a can implement the selected neural network model. After implementing the neural network model, in some instances, UE 115-a can perform postprocessing on the received signal 220 according to the operation sequence indicated in the execution sequence indication 215 for the selected neural network model, and report the output of the neural network model to a higher layer or to base station 105-a during transmission.

[0135] Alternatively, the network can enhance the neural network model using information for performing signal processing for the neural network model. In this example, configuration message 210 may include the neural network model and signal processing functions for the neural network model (e.g., a set of operations, a sequence of operations to be performed, and input and output parameters for each operation). Base station 105-a may encode and output (e.g., provide, transmit) the neural network model and signal processing functions using a format (e.g., ONNX format), and UE 115-a may obtain (e.g., receive) and decode the neural network model and signal processing functions according to this format. In this case, UE 115-a may not be pre-configured with a set of operations for signal processing and receive execution sequence indication 215; instead, UE 115-a may receive configuration message 210 for a certain neural network model and perform signal processing for the neural network model based on configuration message 210. In some instances, UE 115-a can receive multiple configuration messages 210 to obtain information about signal processing used for multiple neural network models.

[0136] In some instances, base station 105-a can receive signaling transmissions indicating information related to signal processing for neural network models. For example, base station 105-a can receive execution sequence instructions from machine learning blocks (e.g., real-time and non-real-time radio access network (RAN) intelligent controllers) for one or more neural network models. Execution sequence instructions can instruct commands to execute a set of operations pre-configured at base station 105-a for signal processing of one or more neural network models. When base station 105-a receives a signal, it can perform signal processing on the signal based on the execution sequence instructions.

[0137] In another instance, UE 115-a may acquire or receive multiple neural network models, each of which may be effective under different operating ranges. In this instance, the set of operations pre-configured at UE 115-a may include one or more operations for determining one or more characteristics of the received signal. For example, UE 115-a may be pre-configured with operations for calculating SNR. Execution sequence indication 215 may instruct the execution of one or more operations as part of signal processing to determine one or more characteristics of the received signal and to select a neural network model based on the output of one or more operations. For example, execution sequence indication 215 may specify that if the SNR of signal 220 is below a threshold, a first neural network model is selected, and if the SNR is above a threshold, a second neural network model is selected. One or more operations may be instances of tables or functions that take some inputs (e.g., the SNR of signal 220) and produce a model identifier (ID) as output.

[0138] In some instances, UE 115-a may indicate its ability to support a pre-configured set of operations at UE 115-a. For example, UE 115-a may output or transmit a signal to base station 105-a indicating that it supports basic mathematics but not complex operations (e.g., singular value decomposition (SVD) or QR decomposition). In another instance, UE 115-a may indicate the threshold input dimension for one or more operations in the pre-configured set of operations at UE 115-a. For example, UE 115-a may output or transmit a signal to base station 105-a indicating that UE 115-a supports 4x4 SVD but not 8x8 SVD or 16x8 SVD. Furthermore, UE 115-a may indicate the threshold runtime for one or more operations in the pre-configured set of operations at UE 115-a. If UE 115-a indicates that it cannot support one or more operations in the operation set, base station 105-a may output or transmit additional network models to UE 115-a.

[0139] Figure 3 illustrates an example of a flowchart 300, according to one or more versions of the present invention, supporting techniques for instructing signal processing procedures for deploying neural network models on a network. In some instances, flowchart 300 may implement various versions of wireless communication system 100 and wireless communication system 200. For example, flowchart 300 may be implemented by UE 115 or base station 105, as described with reference to Figures 1 and 2.

[0140] In some instances, the wireless device may be pre-configured with a set of operations and receive / receive signal transmissions indicating operation sequences (e.g., operation sequence 330) or output operation sequences (e.g., operation sequence 335) for input processing of each neural network model provided to the wireless device. Additionally, the signal transmissions may indicate input and output parameters for each operation of the operation sequence.

[0141] In some instances, the wireless device may acquire or receive signals (e.g., from a base station) at 305 via one or more antennas 301-a. In some instances, the wireless device may determine the characteristics of the signal and select a neural network model based on those characteristics. For example, the wireless device may determine that the signal is associated with an SNR above a threshold and select a first neural network. Alternatively, the wireless device may determine that the signal is associated with an SNR below a threshold and select a second neural network. Once the wireless device identifies which neural network model to implement, it may perform input processing on the received signal at 310. In some instances, the wireless device may perform input processing according to an operation sequence 330. In one instance, the wireless device may implement a neural network model for LOS delay detection. In this case, the operation sequence 330 may be as follows: CFR operation, zero-padding operation, IFFT operation, signal scaling operation, peak search operation, cyclic shift operation, truncation operation, concatenation operation, and realignment operation, and each operation may have specified output and input parameters. Input processing can convert the received signal into a format compatible with the selected neural network model, and thus, the wireless device can implement the selected neural network model at 315.

[0142] At 320, the wireless device can perform output processing. In some instances, the wireless device can perform output processing according to operation sequence 335. In one instance, the wireless device can implement a neural network model for LOS delay detection. In this instance, operation sequence 335 can instruct the output to be shifted and scaled via the result of a cyclic shift operation performed at 310. In some instances, operation sequence 335 and operation sequence 330 can include different operations or one or more of the same operations. After performing output processing, the wireless device can map the output of the neural network model to one or more reports at 325, and the wireless device can send the reports to a higher layer 340, or output / transmit the reports to the network or base station via one or more antennas 301-b.

[0143] Figure 4 illustrates an example of a machine learning process 400, according to one or more forms of the present invention, supporting techniques for instructing signal processing procedures for deploying neural network models on a network. The machine learning process 400 can be implemented at a wireless device (such as UE 115 as described with reference to Figures 1-3). The machine learning process 400 may include machine learning algorithms 410. In some instances, the wireless device may receive a neural network model from base station 105 and implement one or more machine learning algorithms 410 as part of the neural network model to optimize the communication process.

[0144] As shown in the figure, the machine learning algorithm 410 can be an instance of a neural network, such as a feedforward (FF) or deep feedforward (DFF) neural network, a recursive neural network (RNN), a long / short-term memory (LSTM) neural network, or any other type of neural network. However, UE 115 can support any other machine learning algorithm. For example, machine learning algorithm 410 can implement the nearest neighbor algorithm, the linear regression algorithm, the simple Bayesian algorithm, the random forest algorithm, or any other machine learning algorithm. Furthermore, the machine learning process 400 may involve supervised learning, unsupervised learning, semi-supervised learning, reinforcement learning, or any combination thereof. Machine learning can be implemented before the deployment of UE 115, during the deployment of UE 115, during low usage periods of UE 115 at the time of deployment, or any combination thereof.

[0145] The machine learning algorithm 410 may include an input layer 415, one or more hidden layers 420, and an output layer 425. In a fully connected neural network with one hidden layer 420, each hidden layer node 435 may receive values ​​as input from each input layer node 430, where each input is weighted. These neural network weights may be based on a cost function modified during the training of the machine learning algorithm 410. Similarly, each output layer node 440 may receive values ​​as input from each hidden layer node 435, where the inputs are weighted. If post-deployment training (e.g., online training) is supported at the UE 115, the UE 115 may allocate memory to store errors and / or gradients used for inverse matrix multiplication. These errors and / or gradients may support updating the machine learning algorithm 410 based on output feedback. Training the machine learning algorithm 410 can support the computation of weights (e.g., connecting input layer node 430 to hidden layer node 435, and connecting hidden layer node 435 to output layer node 440) to map input patterns to desired output results. This training may result in a UE-specific machine learning algorithm 410 based on historical application data and data transmission for a particular UE 115.

[0146] The UE 115 can send the input value 405 to the machine learning algorithm 410 for processing. In some instances, the UE 115 can preprocess the input value 405 according to the operation sequence received from the base station, so that the input value 405 can have a format compatible with the machine learning algorithm 410. The input value 405 can be converted into a set of k input layer nodes 430 at the input layer 415. In some cases, different measurements can be input at different input layer nodes 430 of the input layer 415. If the number of input layer nodes 430 exceeds the number of inputs corresponding to the input value 405, preset values ​​(e.g., values ​​of 0) can be assigned to some input layer nodes 430. As shown, the input layer 415 may include three input layer nodes 430-a, 430-b, and 430-c. However, it should be understood that the input layer 415 may include any number of input layer nodes 430 (e.g., 20 input nodes).

[0147] Machine learning algorithm 410 can transform input layer 415 into hidden layer 420 based on several input-to-hidden weights between k input layer nodes 430 and n hidden layer nodes 435. Machine learning algorithm 410 may include any number of hidden layers 420 as an intermediate step between input layer 415 and output layer 425. Additionally, each hidden layer 420 may include any number of nodes. For example, as shown, hidden layer 420 may include four hidden layer nodes 435-a, 435-b, 435-c, and 435-d. However, it should be understood that hidden layer 420 may include any number of hidden layer nodes 435 (e.g., 10 input nodes). In a fully connected neural network, each node in a layer may be based on each node in the previous layer. For example, the value of hidden layer node 435-a may be based on the values ​​of input layer nodes 430-a, 430-b, and 430-c (e.g., where different weights are applied to each node value).

[0148] The machine learning algorithm 410 can determine the values ​​of output layer nodes 440 for output layer 425 after one or more hidden layers 420. For example, the machine learning algorithm 410 can transform the hidden layer 420 into the output layer 425 based on several hidden-to-output weights between n hidden layer nodes 435 and m output layer nodes 440. In some cases, n=m. Each output layer node 440 can correspond to a different output value 445 of the machine learning algorithm 410. As shown, the machine learning algorithm 410 can include three output layer nodes 440-a, 440-b, and 440-c, thereby supporting three different thresholds. However, it should be understood that the output layer 425 can include any number of output layer nodes 440. In some instances, the UE 115 can perform post-processing on the output value 445 according to the operation sequence received from the base station, such that the input value 405 can have a format compatible with reporting the output value 445 to a higher layer or reporting it to the base station 105 during transmission.

[0149] Figure 5 illustrates an example of a process flow 500 supporting one or more states of a technique for instructing a signal processing procedure for deploying a neural network model according to the content of this invention. In some instances, process flow 500 may implement or be implemented by various states of wireless communication system 100, wireless communication system 200, and flowchart 300. Process flow 500 may involve UE 115-b receiving a signal transmission for instructing information related to signal transmission processing for one or more neural network models. Alternative instances may be implemented, in which some steps are performed in a different order than described, or not at all. In some cases, steps may include other features not mentioned below, or additional steps may be added.

[0150] At 505, UE 115-b may potentially output or transmit capability information to base station 105-b. The capability information may indicate the ability of UE 115-b to support one or more signal processing operations. In some instances, the capability information may include an indication of a threshold input dimension for one or more signal processing operations or a threshold runtime for each of one or more signal processing operations.

[0151] At 510, UE 115-b can obtain or receive configuration information from base station 105-b. The configuration information may include indications of one or more neural network models. In some instances, base station 105-b may determine which one or more neural network models to output / transmit to UE 115-b based on current operating conditions (e.g., number of antennas, operating SNR, operating bandwidth portion, modulation or RF model) or based on capability information received at 505.

[0152] At 515, for each of the neural network modes of one or more neural network models provided to UE 115-b at 510, UE 115-b may obtain or receive instructions for an operation sequence from base station 105-b. In one instance, UE 115-b may be configured with a set of operations associated with signal processing (e.g., basic functions or non-trainable layers), and the operation sequence may specify the order in which at least a subset of the set of operations for signal processing (preprocessing or postprocessing) is performed. The operation sequence may also include input and output parameters for each operation in the subset. In another instance, the set of operations and the operation sequence for the neural network of one or more neural network models may be included in the configuration message received at 510. In this case, at 515, UE 115-b may not receive instructions for the operation sequence.

[0153] At 520, UE 115-b can obtain or receive signals from base station 105-b. In some instances, UE 115-b can determine the characteristics of the signal (e.g., SNR, bandwidth, or signal scaling) and select the neural network model to be implemented based on these characteristics. In some cases, UE 115-b can be pre-configured with a table or function that indicates the relationship between the neural network model and the operating range (e.g., SNR range, bandwidth range, or signal scaling range), and UE 115-b can select the neural network model based on this table or function as part of signal processing (e.g., preprocessing).

[0154] At 525, UE 115-b can perform input processing on the signal received at 520. In some instances, UE 115-b can perform input processing according to the operation sequence indicated at 515 or the operation sequence included in the configuration message received at 510.

[0155] At 530, UE 115-b can apply a neural network model.

[0156] At 535, UE 115-b can perform output processing on the neural network model output. In some instances, UE 115-b can perform output processing according to the operation sequence indicated at 515 or included in the configuration message received at 510. After performing output processing, UE 115-b can map the output to one or more reports, and potentially output or transmit one or more reports to base station 105-b.

[0157] Figure 6 illustrates a block diagram 600 of a device 605 supporting one or more states of a signal processing procedure for deploying a neural network model according to the present invention. Device 605 may be an example of a UE 115 as described herein. Device 605 may include a receiver 610, a transmitter 615, and a communication manager 620. Device 605 may also include a processor. Each of these elements may communicate with each other (e.g., via one or more buses).

[0158] Receiver 610 may provide components for acquiring (e.g., receiving) information (such as packets, user data, control information, or any combination thereof) associated with various information channels (e.g., control channels, data channels, information channels associated with techniques for instructing signal processing procedures for deploying neural network models on a network). Information may be transmitted to other elements of device 605. Receiver 610 may utilize a single antenna or a collection of multiple antennas.

[0159] Transmitter 615 may provide components for outputting (e.g., providing, transmitting) signals generated by other elements of device 605. For example, transmitter 615 may transmit information (such as packets, user data, control information, or any combination thereof) associated with various information channels (e.g., control channels, data channels, information channels associated with techniques for instructing signal processing procedures for deploying neural network models on a network). In some instances, transmitter 615 may be co-located with receiver 610 in a transceiver module. Transmitter 615 may utilize a single antenna or a collection of multiple antennas.

[0160] The communication manager 620, receiver 610, transmitter 615, or various combinations thereof or various elements thereof may be instances of various types of components used to perform the techniques described herein for instructing signal processing procedures for deploying neural network models on a network. For example, the communication manager 620, receiver 610, transmitter 615, or various combinations thereof or elements thereof may support methods for performing one or more of the functions described herein.

[0161] In some instances, the communication manager 620, receiver 610, transmitter 615, or various combinations or elements thereof may be implemented in hardware (e.g., in a communication management circuitry system). The hardware may include processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, individual gate or transistor logic, individual hardware elements, or any combination thereof, configured to or otherwise support components for performing the functions described herein. In some instances, the processor and memory coupled to the processor may be configured to perform one or more of the functions described herein (e.g., via instructions stored in memory executed by the processor).

[0162] Alternatively or in some instances, the communication manager 620, receiver 610, transmitter 615, or various combinations or elements 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 620, receiver 610, transmitter 615, or various combinations or elements thereof may be performed by a general-purpose processor, DSP, central processing unit (CPU), ASIC, FPGA, or any combination of such or other programmable logic devices (e.g., components configured or otherwise supported for performing the functions described herein).

[0163] In some instances, the communication manager 620 may be configured to use or otherwise cooperate with the receiver 610, transmitter 615, or both to perform various operations (e.g., acquire / receive, monitor, output / transmit). For example, the communication manager 620 may receive information from the receiver 610, send information to the transmitter 615, or integrate with the receiver 610, transmitter 615, or both to receive information, transmit information, or perform various other operations as described herein.

[0164] According to the examples disclosed herein, the communication manager 620 can support wireless communication at a device in a wireless network. For example, the communication manager 620 can be configured or otherwise supported to support means for receiving configuration messages for a device, the configuration messages indicating one or more neural network models for the device. The communication manager 620 can be configured or otherwise supported to support means for receiving instructions on an operation sequence for a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure or an output preprocessing procedure associated with at least one neural network model. The communication manager 620 can be configured or otherwise supported to support means for performing a signal processing procedure for at least one neural network model using signals received at the device, according to the operation sequence.

[0165] By including or configuring the communication manager 620 according to the examples described herein, device 605 (e.g., a processor that controls or is otherwise coupled to receiver 610, transmitter 615, communication manager 620, or a combination thereof) can support techniques for reducing processing and power consumption. Receiving information related to signal processing can allow device 605 to implement a neural network model. By implementing a neural network model, device 605 can optimize the communication process, which in turn can reduce power consumption at device 605.

[0166] Figure 7 illustrates a block diagram 700 of a device 705, according to one or more states of the present invention, supporting techniques for instructing signal processing procedures for deploying neural network models on a network. Device 705 may be an example of a state of device 605 or 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 elements may communicate with each other (e.g., via one or more buses).

[0167] Receiver 710 may provide components for acquiring (e.g., receiving) information (such as packets, user data, control information, or any combination thereof) associated with various information channels (e.g., control channels, data channels, information channels associated with technologies used to instruct signal processing procedures for deploying neural network models on a network). Information may be transmitted to other elements of device 705. Receiver 710 may utilize a single antenna or a collection of multiple antennas.

[0168] Transmitter 715 may provide components for outputting (e.g., transmitting) signals generated by other elements of device 705. For example, transmitter 715 may transmit information (such as packets, user data, control information, or any combination thereof) associated with various information channels (e.g., control channels, data channels, information channels associated with techniques for instructing signal processing procedures for deploying neural network models on a network). In some instances, transmitter 715 may be co-located with receiver 710 in a transceiver module. Transmitter 715 may utilize a single antenna or a collection of multiple antennas.

[0169] Device 705 or its various elements may be instances of components for performing various forms of techniques as described herein for instructing signal processing procedures for deploying neural network models on a network. For example, communication manager 720 may include UE model manager 725, UE signal processing manager 730, execution element 735, or any combination thereof. Communication manager 720 may be an instance of a form of communication manager 620 as described herein. In some instances, communication manager 720 or its various elements may be configured to use receiver 710, transmitter 715, or both, or otherwise cooperate with receiver 710, transmitter 715, or both to perform various operations (e.g., acquire / receive, monitor, output / transmit). For example, communication manager 720 may receive information from receiver 710, send information to transmitter 715, or be integrated with receiver 710, transmitter 715, or both to receive information, transmit information, or perform various other operations as described herein.

[0170] According to the examples disclosed herein, the communication manager 720 can support wireless communication at a device in a wireless network. The UE model manager 725 can be configured or otherwise supported to support components for receiving configuration messages for a device, the configuration messages indicating one or more neural network models for the device. The UE signal processing manager 730 can be configured or otherwise supported to support components for receiving instructions on an operation sequence for a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure associated with at least one neural network model or an output preprocessing procedure associated with at least one neural network model. The execution element 735 can be configured or otherwise supported to support components for performing a signal processing procedure for at least one neural network model using signals received at the device, according to the operation sequence.

[0171] Figure 8 illustrates a block diagram 800 of a communication manager 820 according to one or more forms of the present invention, supporting techniques for instructing signal processing procedures for deploying neural network models on a network. The communication manager 820 may be an instance of a communication manager 620, communication manager 720, or both as described herein. The communication manager 820 or its various elements may be instances of components for executing various forms of techniques for instructing signal processing procedures for deploying neural network models on a network, as described herein. For example, the communication manager 820 may include a UE model manager 825, a UE signal processing manager 830, an execution element 835, a UE capability manager 840, or any combination thereof. Each of these elements may communicate directly or indirectly with each other (e.g., via one or more buses).

[0172] According to the examples disclosed herein, the communication manager 820 can support wireless communication at a device in a wireless network. The UE model manager 825 can be configured or otherwise supported to support components for receiving configuration messages for a device, the configuration messages indicating one or more neural network models for the device. The UE signal processing manager 830 can be configured or otherwise supported to support components for receiving instructions on an operation sequence for a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure associated with at least one neural network model or an output preprocessing procedure associated with at least one neural network model. The execution element 835 can be configured or otherwise supported to support components for performing a signal processing procedure for at least one neural network model using signals received at the device, according to the operation sequence.

[0173] In some instances, the UE signal processing manager 830 may be configured or otherwise supported to receive components that configure the device with a set of operations, the set of operations including one or more operations for a sequence of operations for at least one neural network model. In some instances, the device may include a UE, a base station, a network entity, a relay device, a sidelink device, or an IAB node.

[0174] In some instances, the UE signal processing manager 830 may be configured or otherwise support a component for receiving an instruction in a configuration message for an operation sequence for a signal processing procedure, wherein the configuration message includes an operation set that includes all operations for an operation sequence for at least one neural network model.

[0175] In some instances, the UE signal processing manager 830 may be configured or otherwise support a component for receiving a second configuration message for the device, the second configuration message indicating: at least one neural network model, an indication of an operation sequence, and an operation set including all operations for the operation sequence of at least one neural network model.

[0176] In some instances, the UE signal processing manager 830 may be configured or otherwise support a component for receiving a set of input parameters, a set of output parameters, or both of an operation sequence for at least one neural network model.

[0177] In some instances, the UE model manager 825 may be configured or otherwise supported to receive an instruction for a mapping between one or more neural network models and a set of operating conditions, wherein the execution of a signal processing procedure for at least one neural network model using signals received at the device according to an operating sequence is based on a mapping between one or more neural network models and a set of operating conditions.

[0178] In some instances, the set of operating conditions includes a signal-to-noise ratio range, bandwidth range, signal scaling range, channel delay profile, signal peak range, or any combination thereof associated with the signal received at the device.

[0179] In some instances, the UE capability manager 840 may be configured or otherwise supported to include components for transmitting messages indicating the device's capability to support one or more operations for one or more signal processing procedures, wherein receiving an indication of an operation sequence for a signal processing procedure for at least one neural network model is based on the device's capabilities. In some instances, the operation sequence includes one or more operations supported by the device.

[0180] In some instances, the message indicating the device’s ability to support one or more operations for signal processing includes: an indication of a threshold input dimension for each of the one or more operations for signal processing or a threshold runtime for each of the one or more operations for signal processing.

[0181] In some instances, the UE signal processing manager 830 may be configured or otherwise supported for receiving RRC signaling transmissions or MAC-CEs, including indications of operation sequences.

[0182] In some instances, the UE signal processing manager 830 may be configured or otherwise supported to receive instructions for one or more data formats associated with one or more operations in a sequence of operations, including XML data format, JSON data format, or any combination thereof.

[0183] Figure 9 illustrates a system 900 including a device 905 supporting technology for instructing a neural network model for network deployment, according to one or more embodiments of the present invention. Device 905 may be an example of device 605, device 705, or UE 115 as described herein, or may include elements thereof. Device 905 may wirelessly communicate with one or more network entities 105, UE 115, or any combination thereof. Device 905 may include elements for bidirectional voice and data communication, including elements for transmitting and receiving communications, such as a communication manager 920, I / O controller 910, transceiver 915, antenna 925, memory 930, code 935, and processor 940. These elements may be electronically communicated via one or more buses (e.g., bus 945) or otherwise coupled (e.g., operational ground, communication ground, functional ground, electronic ground, electrical ground).

[0184] The I / O controller 910 can manage input and output signals to the device 905. The I / O controller 910 can also manage peripheral devices not integrated into the device 905. In some cases, the I / O controller 910 can represent a physical connection or port to an external peripheral device. In some cases, the I / O controller 910 can utilize an operating system such as iOS®, ANDROID®, MS-DOS®, MS-WINDOWS®, OS / 2®, UNIX®, LINUX®, or another known operating system. Alternatively or concurrently, the I / O controller 910 can represent or interact with a modem, keyboard, mouse, touchscreen, or similar device. In some cases, the I / O controller 910 can be implemented as part of a processor (such as processor 940). In some cases, a user can interact with the device 905 via the I / O controller 910 or via hardware components controlled by the I / O controller 910.

[0185] In some cases, device 905 may include a single antenna 925. However, in other cases, device 905 may have more than one antenna 925, capable of transmitting or receiving multiple wireless transmissions simultaneously. Transceiver 915 may communicate bidirectionally via one or more antennas 925, wired or wireless links as described herein. For example, transceiver 915 may represent a wireless transceiver and may communicate bidirectionally with another wireless transceiver. Transceiver 915 may also include a modem for modulating packets, providing modulated packets to one or more antennas 925 for transmission, and demodulating packets received from one or more antennas 925. Transceiver 915, or transceiver 915 and one or more antennas 925, may be an example of transmitter 615, transmitter 715, receiver 610, receiver 710, or any combination thereof or elements thereof as described herein.

[0186] Memory 930 may include random access memory (RAM) and read-only memory (ROM). Memory 930 may store computer-readable, computer-executable code 935, which includes instructions that, when executed by processor 940, cause device 905 to perform the various functions described herein. Code 935 may be stored in a non-transitory computer-readable medium (such as system memory or other types of memory). In some cases, code 935 may not be directly executable by processor 940, but may cause a computer (e.g., when compiled and executed) to perform the functions described herein. In some cases, in addition, memory 930 may also include a basic I / O system (BIOS) that controls basic hardware or software operations, such as interaction with peripheral components or devices.

[0187] Processor 940 may include intelligent hardware devices (e.g., general-purpose processors, DSPs, CPUs, microcontrollers, ASICs, FPGAs, programmable logic devices, individual gate or transistor logic elements, individual hardware elements, or any combination thereof). In some cases, processor 940 may be configured to operate a memory array using a memory controller. In other cases, the memory controller may be integrated into processor 940. Processor 940 may be configured to execute computer-readable instructions stored in memory (e.g., memory 930) to cause device 905 to perform various functions (e.g., functions or tasks supporting techniques for instructing signal processing procedures for deploying neural network models on a network). For example, device 905 or elements thereof may include processor 940 and memory 930 coupled to processor 940, processor 940 and memory 930 being configured to perform the various functions described herein.

[0188] According to the examples disclosed herein, the communication manager 920 can support wireless communication at a device in a wireless network. For example, the communication manager 920 can be configured or otherwise supported to support means for receiving configuration messages for a device, the configuration messages indicating one or more neural network models for the device. The communication manager 920 can be configured or otherwise supported to support means for receiving instructions on an operation sequence for a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure or an output preprocessing procedure associated with at least one neural network model. The communication manager 920 can be configured or otherwise supported to support means for performing a signal processing procedure for at least one neural network model using signals received at the device according to the operation sequence.

[0189] By including or configuring the communication manager 920 according to the examples described herein, the device 905 can support techniques for improved user experience related to reduced processing and lower power consumption. The methods described herein can support the deployment of new neural network models that can improve performance compared to existing neural network models.

[0190] In some instances, the communication manager 920 may be configured to perform various operations (e.g., receiving, monitoring, transmitting) using or in cooperation with the transceiver 915, one or more antennas 925, or any combination thereof. Although the communication manager 920 is shown as a separate element, in some instances, one or more functions described with reference to the communication manager 920 may be supported or executed by the processor 940, memory 930, code 935, or any combination thereof. For example, code 935 may include instructions executable by the processor 940 to cause the device 905 to execute various forms of techniques as described herein for instructing signal processing procedures for deploying neural network models on a network, or the processor 940 and memory 930 may be otherwise configured to perform or support such operations.

[0191] Figure 10 illustrates a block diagram 1000 of a device 1005, according to one or more configurations of the present invention, supporting techniques for instructing signal processing procedures for deploying neural network models on a network. Device 1005 may be an example of a base station 105 as described herein. Device 1005 may include a receiver 1010, a transmitter 1015, and a communication manager 1020. Device 1005 may also include a processor. Each of these components may communicate with each other (e.g., via one or more buses).

[0192] Receiver 1010 may provide components for receiving information (such as packets, user data, control information, or any combination thereof) associated with various information channels (e.g., control channels, data channels, information channels, control channels, etc., related to technologies used to instruct signal processing procedures for deploying neural network models on a network). Information may be transmitted to other components of device 1005. Receiver 1010 may utilize a single antenna or a collection of multiple antennas.

[0193] Transmitter 1015 may provide components for transmitting signals generated by other elements of device 1005. For example, transmitter 1015 may transmit information (such as packets, user data, control information, or any combination thereof) associated with various information channels (e.g., control channels, data channels, information channels associated with techniques for instructing signal processing procedures for deploying neural network models on a network). In some instances, transmitter 1015 may be co-located with receiver 1010 in a transceiver module. Transmitter 1015 may utilize a single antenna or a collection of multiple antennas.

[0194] The communication manager 1020, receiver 1010, transmitter 1015, or various combinations thereof, or various elements thereof, may be instances of various types of components used to perform the techniques described herein for instructing signal processing procedures for deploying neural network models on a network. For example, the communication manager 1020, receiver 1010, transmitter 1015, or various combinations thereof, or elements thereof, may support methods for performing one or more of the functions described herein.

[0195] In some instances, the communication manager 1020, receiver 1010, transmitter 1015, or various combinations or elements thereof may be implemented in hardware (e.g., in a communication management circuitry system). The hardware may include processors, DSPs, ASICs, FPGAs, or other programmable logic devices, individual gate or transistor logic, individual hardware elements, or any combination thereof configured to or otherwise support components for performing the functions described herein. In some instances, the processor and memory coupled to the processor may be configured to perform one or more of the functions described herein (e.g., via instructions stored in memory executed by the processor).

[0196] Alternatively or in some instances, the communication manager 1020, receiver 1010, transmitter 1015, or various combinations or elements 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 1020, receiver 1010, transmitter 1015, or various combinations or elements thereof may be performed by a general-purpose processor, DSP, CPU, ASIC, FPGA, or any combination of such or other programmable logic devices (e.g., components configured or otherwise supported for performing the functions described herein).

[0197] In some instances, the communication manager 1020 may be configured to use the receiver 1010, the transmitter 1015, or both, or otherwise cooperate with the receiver 1010, the transmitter 1015, or both to perform various operations (e.g., receiving, monitoring, transmitting). For example, the communication manager 1020 may receive information from the receiver 1010, send information to the transmitter 1015, or combine with the receiver 1010, the transmitter 1015, or both to receive information, transmit information, or perform various other operations as described herein.

[0198] According to the examples disclosed herein, the communication manager 1020 can support wireless communication at a base station. For example, the communication manager 1020 can be configured or otherwise supported to support means for transmitting configuration messages to a device, the configuration messages indicating one or more neural network models for the device. The communication manager 1020 can be configured or otherwise supported to support means for transmitting instructions on an operational sequence of a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure associated with at least one neural network model or an output preprocessing procedure associated with at least one neural network model. The communication manager 1020 can be configured or otherwise supported to support means for transmitting signals to a device based on instructions on transmitting an operational sequence of a signal processing procedure for at least one neural network model.

[0199] By including or configuring the communication manager 1020 according to the examples described herein, the device 1005 (e.g., a processor that controls or is otherwise coupled to the receiver 1010, transmitter 1015, communication manager 1020, or a combination thereof) can support techniques for reducing processing and power consumption.

[0200] Figure 11 illustrates a block diagram 1100 of a device 1105, according to one or more configurations of the present invention, supporting techniques for instructing signal processing procedures for deploying neural network models on a network. Device 1105 may be an example of a configuration of device 1005 or base station 105 as described herein. Device 1105 may include a receiver 1110, a transmitter 1115, and a 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).

[0201] Receiver 1110 may provide components for receiving information (such as packets, user data, control information, or any combination thereof) associated with various information channels (e.g., control channels, data channels, information channels, control channels, etc., related to technologies used to instruct signal processing procedures for deploying neural network models on a network). Information may be transmitted to other components of device 1105. Receiver 1110 may utilize a single antenna or a collection of multiple antennas.

[0202] Transmitter 1115 may provide components for transmitting signals generated by other elements of device 1105. For example, transmitter 1115 may transmit information (such as packets, user data, control information, or any combination thereof) associated with various information channels (e.g., control channels, data channels, information channels associated with techniques for instructing signal processing procedures for deploying neural network models on a network). In some instances, transmitter 1115 may be co-located with receiver 1110 in a transceiver module. Transmitter 1115 may utilize a single antenna or a collection of multiple antennas.

[0203] Device 1105 or its various elements may be instances of various types of components used to perform the techniques described herein for instructing signal processing procedures for deploying neural network models on a network. For example, communication manager 1120 may include model manager 1125, signal processing manager 1130, signal transmitter 1135, or any combination thereof. Communication manager 1120 may be instances of various types of communication manager 1020 as described herein. In some instances, communication manager 1120 or its various elements may be configured to use receiver 1110, transmitter 1115, or both, or otherwise cooperate with receiver 1110, transmitter 1115, or both to perform various operations (e.g., receiving, monitoring, transmitting). For example, communication manager 1120 may receive information from receiver 1110, send information to transmitter 1115, or be combined and integrated with receiver 1110, transmitter 1115, or both to receive information, transmit information, or perform various other operations as described herein.

[0204] According to the examples disclosed herein, the communication manager 1120 can support wireless communication at a base station. The model manager 1125 can be configured or otherwise supported to support components for transmitting configuration messages to a device, the configuration messages indicating one or more neural network models for the device. The signal processing manager 1130 can be configured or otherwise supported to support components for transmitting instructions on the sequence of operations for a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure associated with at least one neural network model or an output preprocessing procedure associated with at least one neural network model. The signal transmitter 1135 can be configured or otherwise supported to support components for transmitting signals to a device based on instructions on transmitting the sequence of operations for a signal processing procedure for at least one neural network model.

[0205] Figure 12 illustrates a block diagram 1200 of a communication manager 1220, according to one or more versions of the present invention, supporting techniques for instructing signal processing procedures for deploying neural network models on a network. The communication manager 1220 may be an instance of a communication manager 1020, communication manager 1120, or both as described herein. The communication manager 1220 or its various elements may be instances of components for performing various versions of techniques for instructing signal processing procedures for deploying neural network models on a network, as described herein. For example, the communication manager 1220 may include a model manager 1225, a signal processing manager 1230, a signal transmitter 1235, a capability manager 1240, or any combination thereof. Each of these elements may communicate directly or indirectly with each other (e.g., via one or more buses).

[0206] According to the examples disclosed herein, the communication manager 1220 can support wireless communication at a base station. The model manager 1225 can be configured or otherwise supported to support components for transmitting configuration messages to a device, the configuration messages indicating one or more neural network models for the device. The signal processing manager 1230 can be configured or otherwise supported to support components for transmitting instructions on the sequence of operations of a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure associated with at least one neural network model or an output preprocessing procedure associated with at least one neural network model. The signal transmitter 1235 can be configured or otherwise supported to support components for transmitting signals to a device based on instructions on transmitting the sequence of operations of a signal processing procedure for at least one neural network model.

[0207] In some instances, the signal processing manager 1230 may be configured or otherwise support components for transmitting signals that configure the device with a set of operations, the set of operations including one or more operations for a sequence of operations for at least one neural network model. In some instances, the device may include a UE, a base station, a network entity, a relay device, a sidelink device, or an IAB node.

[0208] In some instances, the signal processing manager 1230 may be configured or otherwise supported to receive a second sequence of operations for a second signaling procedure to be executed at a base station.

[0209] In some instances, the signal processing manager 1230 may be configured or otherwise support components for transmitting instructions for a sequence of operations for a signal processing procedure in a configuration message, wherein the configuration message includes a set of operations that includes all operations for a sequence of operations for at least one neural network model.

[0210] In some instances, the signal processing manager 1230 may be configured or otherwise support components for transmitting a second configuration message to the device, the second configuration message indicating: at least one neural network model, an indication of an operation sequence, and an operation set including all operations for the operation sequence of at least one neural network model.

[0211] In some instances, the signal processing manager 1230 may be configured or otherwise support components for transmitting a set of input parameters, a set of output parameters, or both of one or more operations for transmitting a sequence of operations for at least one neural network model.

[0212] In some instances, the model manager 1225 may be configured or otherwise supported to provide a component for transmitting instructions on the mapping between one or more neural network models and sets of operating conditions.

[0213] In some instances, the set of operating conditions includes a signal-to-noise ratio range, bandwidth range, signal scaling range, channel delay profile, signal peak range, or any combination thereof associated with the signal transmitted to the device.

[0214] In some instances, the capability manager 1240 may be configured or otherwise supported to receive a message indicating that the device supports the capability for one or more operations for one or more signal processing procedures, wherein receiving the indication of the sequence of operations for a signal processing procedure for at least one neural network model is based on the capability of the device.

[0215] In some instances, the sequence of operations includes one or more operations supported by the device.

[0216] In some instances, the message indicating the device’s ability to support one or more operations for signal processing includes: an indication of a threshold input dimension for each of the one or more operations for signal processing or a threshold runtime for each of the one or more operations for signal processing.

[0217] In some instances, the signal processing manager 1230 may be configured or otherwise support components for transmitting RRC signaling or MAC-CE, including indications of operation sequences.

[0218] In some instances, the signal processing manager 1230 may be configured or otherwise support components for transmitting instructions on one or more data formats associated with one or more operations of an operation sequence, including XML data format, JSON data format, or any combination thereof.

[0219] Figure 13 illustrates a system 1300 including a device 1305 supporting technology for instructing a neural network model for network deployment, according to one or more embodiments of the present invention. Device 1305 may be an instance of device 1005, device 1105, or base station 105 as described herein, or include elements thereof. Device 1305 may wirelessly communicate with one or more network entities 105, UE 115, or any combination thereof. Device 1305 may include elements for bidirectional voice and data communication, including elements for transmitting and receiving communications, such as a communication manager 1320, a network communication manager 1310, a transceiver 1315, an antenna 1325, a memory 1330, a code 1335, a processor 1340, and an inter-station communication manager 1345. These components may be electronically communicated or otherwise coupled via one or more buses (e.g., bus 1350).

[0220] The network communication manager 1310 can manage communication with the core network 130 (e.g., via one or more wired backhaul links). For example, the network communication manager 1310 can manage the transmission of data communications to client devices (e.g., one or more UEs 115).

[0221] In some cases, device 1305 may include a single antenna 1325. However, in other cases, device 1305 may have more than one antenna 1325, which is capable of transmitting or receiving multiple wireless transmissions simultaneously. Transceiver 1315 may communicate bidirectionally via one or more antennas 1325, wired or wireless links as described herein. For example, transceiver 1315 may represent a wireless transceiver and may communicate bidirectionally with another wireless transceiver. Transceiver 1315 may also include a modem for modulating packets, providing modulated packets to one or more antennas 1325 for transmission, and demodulating packets received from one or more antennas 1325. Transceiver 1315, or transceiver 1315 and one or more antennas 1325, may be an example of transmitter 1015, transmitter 1115, receiver 1010, receiver 1010, or any combination thereof or elements thereof as described herein.

[0222] Memory 1330 may include RAM and ROM. Memory 1330 may store computer-readable, computer-executable code 1335, which includes instructions that, when executed by processor 1340, cause device 1305 to perform the various functions described herein. Code 1335 may be stored in a non-transitory computer-readable medium (such as system memory or other types of memory). In some cases, code 1335 may not be directly executable by processor 1340, but may cause a computer (e.g., when compiled and executed) to perform the functions described herein. In some cases, in addition, memory 1330 may also include a BIOS, which controls basic hardware or software operations, such as interaction with peripheral components or devices.

[0223] Processor 1340 may include intelligent hardware devices (e.g., general-purpose processors, DSPs, CPUs, microcontrollers, ASICs, FPGAs, programmable logic devices, individual gate or transistor logic elements, individual hardware elements, or any combination thereof). In some cases, processor 1340 may be configured to use a memory controller to operate a memory array. In other cases, the memory controller may be integrated into processor 1340. Processor 1340 may be configured to execute computer-readable instructions stored in memory (e.g., memory 1330) to cause device 1305 to perform various functions (e.g., functions or tasks supporting techniques for instructing signal processing procedures for deploying neural network models on a network). For example, device 1305 or elements thereof may include processor 1340 and memory 1330 coupled to processor 1340, processor 1340 and memory 1330 being configured to perform the various functions described herein.

[0224] The inter-site communication manager 1345 can manage communication with other network entities 105 and may include a controller or scheduler for cooperating with other network entities 105 to control communication with the UE 115. For example, the inter-site communication manager 1345 can coordinate the scheduling of transmissions to the UE 115 to implement various interference mitigation techniques such as beamforming or joint transmission. In some instances, the inter-site communication manager 1345 may provide an X2 interface within LTE / LTE-A radio communication network technology to facilitate communication between network entities 105.

[0225] According to the examples disclosed herein, the communication manager 1320 can support wireless communication at a base station. For example, the communication manager 1320 can be configured or otherwise supported to support means for transmitting configuration messages to a device, the configuration messages indicating one or more neural network models for the device. The communication manager 1320 can be configured or otherwise supported to support means for transmitting instructions on an operational sequence of a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure associated with at least one neural network model or an output preprocessing procedure associated with at least one neural network model. The communication manager 1320 can be configured or otherwise supported to support means for transmitting signals to a device based on instructions on transmitting an operational sequence of a signal processing procedure for at least one neural network model.

[0226] By including or configuring the communication manager 1320 according to the examples described herein, the device 1305 can support techniques for improved user experience related to reduced processing and lower power consumption.

[0227] In some instances, the communication manager 1320 may be configured to perform various operations (e.g., receiving, monitoring, transmitting) using or in cooperation with the transceiver 1315, one or more antennas 1325, or any combination thereof. Although the communication manager 1320 is shown as a separate element, in some instances, one or more functions described with reference to the communication manager 1320 may be supported or executed by the processor 1340, memory 1330, code 1335, or any combination thereof. For example, code 1335 may include instructions executable by the processor 1340 to cause the device 1305 to execute various forms of techniques as described herein for instructing signal processing procedures for deploying neural network models on a network, or the processor 1340 and memory 1330 may be otherwise configured to perform or support such operations.

[0228] Figure 14 illustrates a flowchart of a method 1400 that supports one or more states of a signal processing procedure for deploying a neural network model according to the present invention. The operation of method 1400 can be implemented by a UE or its components as described herein; for example, the operation of method 1400 can be performed by a UE 115 as described with reference to Figures 1 to 9. In some instances, the UE can execute a set of instructions to control the functional units of the UE to perform the described functions. Alternatively, the UE can use dedicated hardware to perform various states of the described functions.

[0229] At 1405, the method may include the steps of: obtaining (e.g., receiving) configuration messages for the device, wherein the configuration messages indicate one or more neural network models for the device. The operation of 1405 may be performed according to examples as disclosed herein. In some instances, various forms of the operation of 1405 may be performed by a UE model manager 825 as described with reference to FIG8.

[0230] At 1410, the method may include the step of: obtaining an indication of an operational sequence for a signal processing procedure for at least one of one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure associated with at least one neural network model or an output preprocessing procedure associated with at least one neural network model. The operation of 1410 may be performed according to examples as disclosed herein. In some examples, various forms of the operation of 1410 may be performed by a UE signal processing manager 830 as described with reference to FIG8.

[0231] At 1415, the method may include the step of performing a signal processing procedure for at least one neural network model using signals obtained at the device according to an operation sequence. The operation of 1415 may be performed according to examples as disclosed herein. In some examples, various forms of the operation of 1415 may be performed by an execution element 835 as described with reference to FIG8.

[0232] Figure 15 illustrates a flowchart of a method 1500 that supports one or more states of a signal processing procedure for deploying a neural network model according to the present invention. Operation of method 1500 can be implemented by a UE or its components as described herein. For example, operation of method 1500 can be performed by a UE 115 as described with reference to Figures 1 to 9. In some instances, the UE can execute a set of instructions to control the functional units of the UE to perform the described functions. Alternatively, the UE can use dedicated hardware to perform various states of the described functions.

[0233] At 1505, the method may include the steps of: obtaining (e.g., receiving) configuration messages for the device, wherein the configuration messages indicate one or more neural network models for the device. The operation of 1505 may be performed according to examples as disclosed herein. In some instances, the various forms of operation of 1505 may be performed by a UE model manager 825 as described with reference to FIG8.

[0234] At 1510, the method may optionally include the step of: obtaining a signal transmission configuring the device with an operation set, the operation set including one or more operations for an operation sequence for at least one neural network model. The operation of 1510 may be performed according to examples as disclosed herein. In some instances, the various forms of the operation of 1510 may be performed by a UE signal processing manager 830 as described with reference to FIG8.

[0235] At 1515, the method may include the step of: obtaining an indication of an operational sequence for a signal processing procedure for at least one of one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure associated with at least one neural network model or an output preprocessing procedure associated with at least one neural network model. The operation at 1515 may be performed according to examples as disclosed herein. In some examples, various forms of the operation at 1515 may be performed by a UE signal processing manager 830 as described with reference to FIG8.

[0236] At 1520, the method may include the step of: performing a signal processing procedure for at least one neural network model using signals obtained at the device, according to an operation sequence. The operation of 1520 may be performed according to examples as disclosed herein. In some examples, various forms of the operation of 1520 may be performed by an execution element 835 as described with reference to FIG8.

[0237] Figure 16 illustrates a flowchart of a method 1600, which supports one or more modes of a signal processing procedure for deploying a neural network model according to the present invention. Operation of method 1600 can be implemented by a UE or its components as described herein. For example, operation of method 1600 can be performed by a UE 115 as described with reference to Figures 1 to 9. In some instances, the UE can execute a set of instructions to control the functional units of the UE to perform the described functions. Alternatively, the UE can use dedicated hardware to perform various modes of the described functions.

[0238] At 1605, the method may include the steps of: obtaining (e.g., receiving) configuration messages for the device, wherein the configuration messages indicate one or more neural network models for the device. The operation of 1605 may be performed according to examples as disclosed herein. In some instances, various forms of the operation of 1605 may be performed by a UE model manager 825 as described with reference to FIG8.

[0239] At 1610, the method may include the step of: obtaining an indication of an operational sequence for a signal processing procedure for at least one of one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure associated with at least one neural network model or an output preprocessing procedure associated with at least one neural network model. The operation of 1610 may be performed according to examples as disclosed herein. In some examples, various forms of the operation of 1610 may be performed by a UE signal processing manager 830 as described with reference to FIG8.

[0240] At 1615, the method may optionally include the step of: obtaining an indication of an operation sequence for a signal processing procedure in a configuration message, wherein the configuration message includes an operation set, the operation set including all operations of the operation sequence for at least one neural network model. The operation at 1615 may be performed according to examples as disclosed herein. In some instances, the various forms of the operation at 1615 may be performed by a UE signal processing manager 830 as described with reference to FIG8.

[0241] At 1620, the method may include the step of: performing a signal processing procedure for at least one neural network model using signals obtained at the device, according to an operation sequence. The operation of 1620 may be performed according to examples as disclosed herein. In some instances, various forms of the operation of 1620 may be performed by an execution element 835 as described with reference to FIG8.

[0242] Figure 17 illustrates a flowchart of a method 1700, which supports one or more modes of a signal processing procedure for deploying a neural network model according to the present invention. Operation of method 1700 can be implemented by a base station or its components as described herein. For example, operation of method 1700 can be performed by a base station 105 as described with reference to Figures 1-5 and Figures 10-13. In some instances, the base station can execute an instruction set to control the functional units of the base station to perform the described functions. Alternatively or concurrently, the base station can use dedicated hardware to perform various modes of the described functions.

[0243] At 1705, the method may include the step of: outputting (e.g., transmitting, providing) a configuration message to the device, wherein the configuration message indicates one or more neural network models for the device. The operation of 1705 may be performed according to examples as disclosed herein. In some instances, various forms of operation of 1705 may be performed by a model manager 1225 as described with reference to FIG12.

[0244] At 1710, the method may include the step of: outputting an indication of an operational sequence of a signal processing procedure for at least one of one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure associated with at least one neural network model or an output preprocessing procedure associated with at least one neural network model. The operation of 1710 may be performed according to examples as disclosed herein. In some examples, various forms of the operation of 1710 may be performed by a signal processing manager 1230 as described with reference to FIG12.

[0245] At 1715, the method may include the step of outputting a signal to a device based on an instruction for an operational sequence of a signal processing procedure for at least one neural network model. The operation of 1715 may be performed according to examples disclosed herein. In some instances, various forms of the operation of 1715 may be performed by a signal transmitter 1235 as described with reference to FIG12.

[0246] Figure 18 illustrates a flowchart of a method 1800, which supports one or more states of a signal processing procedure for deploying a neural network model according to the present invention. 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 performed by a base station 105 as described with reference to Figures 1-5 and Figures 10-13. In some instances, the base station can execute an instruction set to control the functional units of the base station to perform the described functions. Alternatively or concurrently, the base station can use dedicated hardware to perform various states of the described functions.

[0247] At 1805, the method may include the step of: outputting (e.g., transmitting, providing) a configuration message to the device, wherein the configuration message indicates one or more neural network models for the device. The operation of 1805 may be performed according to examples as disclosed herein. In some instances, various forms of operation of 1805 may be performed by a model manager 1225 as described with reference to FIG12.

[0248] At 1810, the method may optionally include the step of: outputting a signal transmission configuring the device with a set of operations, the set of operations including one or more operations for a sequence of operations for at least one neural network model. The operations of 1810 may be performed according to examples as disclosed herein. In some instances, the various forms of the operations of 1810 may be performed by a signal processing manager 1230 as described with reference to FIG12.

[0249] At 1815, the method may include the step of: outputting an indication of an operational sequence of a signal processing procedure for at least one of one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure associated with at least one neural network model or an output preprocessing procedure associated with at least one neural network model. The operation at 1815 may be performed according to examples as disclosed herein. In some examples, various forms of the operation at 1815 may be performed by a signal processing manager 1230 as described with reference to FIG12.

[0250] At 1820, the method may include the step of outputting a signal to a device based on an instruction for an operational sequence of a signal processing procedure for at least one neural network model. The operation of 1820 may be performed according to examples disclosed herein. In some instances, various forms of the operation of 1820 may be performed by a signal transmitter 1235 as described with reference to FIG12.

[0251] Figure 19 illustrates a flowchart of a method 1900, which supports one or more modes of a signal processing procedure for deploying a neural network model according to the present invention. Operation of method 1900 can be implemented by a base station or its components as described herein. For example, operation of method 1900 can be performed by a base station 105 as described with reference to Figures 1-5 and Figures 10-13. In some instances, the base station can execute an instruction set to control the functional units of the base station to perform the described functions. Alternatively or concurrently, the base station can use dedicated hardware to perform various modes of the described functions.

[0252] At 1905, the method may include the step of: outputting (e.g., transmitting, providing) a configuration message to the device, wherein the configuration message indicates one or more neural network models for the device. The operation of 1905 can be performed according to examples as disclosed herein. In some instances, various forms of operation of 1905 may be performed by a model manager 1225 as described with reference to FIG12.

[0253] At 1910, the method may include the step of: outputting an indication of an operational sequence of a signal processing procedure for at least one of one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure associated with at least one neural network model or an output preprocessing procedure associated with at least one neural network model. The operation at 1910 may be performed according to examples as disclosed herein. In some examples, various forms of the operation at 1910 may be performed by a signal processing manager 1230 as described with reference to FIG12.

[0254] At 1915, the method may optionally include the step of: outputting an indication of an operation sequence for a signal processing procedure in a configuration message, wherein the configuration message includes an operation set, the operation set including all operations of the operation sequence for at least one neural network model. The operation at 1915 can be performed according to examples as disclosed herein. In some instances, the various forms of the operation at 1915 can be performed by a signal processing manager 1230 as described with reference to FIG12.

[0255] At 1920, the method may include the step of outputting a signal to a device based on an instruction for an operational sequence of a signal processing procedure for at least one neural network model. The operation at 1920 may be performed according to examples disclosed herein. In some examples, various forms of the operation at 1920 may be performed by a signal transmitter 1235 as described with reference to FIG12.

[0256] The following provides a summary of the various aspects of the case:

[0257] Sample 1: A method for wireless communication at a device in a wireless network, comprising the steps of: obtaining configuration information for the device, wherein the configuration information indicates one or more neural network models for the device; obtaining an instruction for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and, according to the operation sequence, performing the signal processing procedure for the at least one neural network model using signals obtained at the device.

[0258] State 2: The method according to State 34 also includes the following steps: obtaining a signal transmission that configures the device with an operation set, the operation set including one or more operations for the operation sequence of the at least one neural network model.

[0259] State 3: The method according to any one of states 34 to 35, wherein obtaining the instruction for the operation sequence includes: obtaining the instruction for the operation sequence for the signal processing procedure in the configuration message, wherein the configuration message includes an operation set, the operation set including all operations of the operation sequence for the at least one neural network model.

[0260] State 4: The method according to any one of states 34 to 36, wherein obtaining the instruction on the operation sequence includes: obtaining a second configuration message for the device, wherein the second configuration message indicates the at least one neural network model, the instruction on the operation sequence, and an operation set including all operations of the operation sequence for the at least one neural network model.

[0261] State 5: The method according to any one of states 34 to 37, wherein obtaining the instruction on the operation sequence includes: obtaining a set of input parameters, a set of output parameters, or both of the operation sequence for the at least one neural network model.

[0262] State 6: The method according to any one of states 34 to 38 also includes the step of: obtaining an indication of a mapping between the one or more neural network models and the set of operating conditions, wherein the signal processing procedure for the at least one neural network model is executed, at least in part, based on the mapping between the one or more neural network models and the set of operating conditions, according to the sequence of operations, using the signal obtained at the device.

[0263] State 7: The method according to State 39, wherein the set of operating conditions includes a signal-to-noise ratio range, bandwidth range, signal scaling range, channel delay profile, signal peak range, or any combination thereof associated with the signal obtained at the device.

[0264] State 8: The method according to any one of states 34 to 40 also includes the step of: outputting a message indicating the capability of the device to support one or more operations for one or more signal processing programs, wherein the indication of the sequence of operations for the signal processing program for the at least one neural network model is obtained at least in part based on the capability of the device.

[0265] State 9: The method according to state 41, wherein the operation sequence includes one or more operations supported by the device.

[0266] State 10: According to any one of states 41 to 42, the message indicating that the device supports the capability for the one or more operations of the one or more signal processing programs includes: an indication of a threshold input dimension for each of the one or more operations of the one or more signal processing programs or a threshold runtime for each of the one or more operations of the one or more signal processing programs.

[0267] State 11: The method according to any one of states 34 to 43, wherein obtaining the indication of the operation sequence includes: obtaining an RRC signal transmission or MAC-CE including the indication of the operation sequence.

[0268] State 12: The method according to any one of states 34 to 44 also includes the step of: obtaining an indication of one or more data formats associated with one or more operations of the operation sequence, wherein the one or more data formats include XML data format, JSON data format, or any combination thereof.

[0269] State 13: The method according to any one of States 34 to 35, wherein the device includes a UE, a base station, a network entity, a relay device, a sidelink device, or an IAB node.

[0270] Sample 14: A method for wireless communication at a network entity, comprising the steps of: outputting a configuration message to a device, wherein the configuration message indicates one or more neural network models for the device; outputting an instruction for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and outputting a signal to the device at least in part based on the instruction for the operation sequence of the signal processing procedure for the at least one neural network model.

[0271] State 15: The method according to State 47 also includes the step of: outputting a signal transmission that configures the device with an operation set, the operation set including one or more operations for the operation sequence of the at least one neural network model.

[0272] State 16: The method according to any one of states 47 to 48 also includes the step of: obtaining a second operation sequence for executing a second signaling procedure at the network entity.

[0273] State 17: The method according to any one of states 47 to 49, wherein outputting the indication of the operation sequence includes: outputting an indication of the operation sequence for the signal processing procedure in the configuration message, wherein the configuration message includes an operation set, the operation set including all operations of the operation sequence for the at least one neural network model.

[0274] State 18: The method according to any one of states 47 to 50, wherein outputting the indication to the operation sequence includes: outputting a second configuration message to the device, wherein the second configuration message indicates: the at least one neural network model, the indication to the operation sequence, and an operation set including all operations for the operation sequence of the at least one neural network model.

[0275] State 19: The method according to any one of states 47 to 51, wherein the instruction on the operation sequence is output includes: outputting a set of input parameters, a set of output parameters, or both of the operation sequence for the at least one neural network model.

[0276] State 20: The method according to any one of states 47 to 52 also includes the step of: outputting an indication of the mapping between the one or more neural network models and the set of operating conditions.

[0277] State 21: According to the method of State 53, wherein the set of operating conditions includes a signal-to-noise ratio range, bandwidth range, signal scaling range, channel delay profile, signal peak range, or any combination thereof associated with the signal transmitted to the device.

[0278] State 22: The method according to any one of states 47 to 54 also includes the step of: obtaining a message indicating that the device supports one or more operations for one or more signal processing programs, wherein obtaining the indication of the sequence of operations for the signal processing program for the at least one neural network model is at least partially based on the capability of the device.

[0279] State 23: According to the method of state 55, wherein the operation sequence includes one or more operations supported by the device.

[0280] State 24: According to any one of states 55 to 56, the message indicating that the device supports the capability for the one or more operations of the one or more signal processing programs includes: an indication of a threshold input dimension for each of the one or more operations of the one or more signal processing programs or a threshold runtime for each of the one or more operations of the one or more signal processing programs.

[0281] State 25: The method according to any one of states 47 to 57, wherein the output of the indication to the operation sequence includes: outputting an RRC signal transmission or MAC-CE including the indication to the operation sequence.

[0282] State 26: The method according to any one of states 47 to 58 also includes the step of: outputting an indication of one or more data formats associated with one or more operations of the operation sequence, wherein the one or more data formats include XML data format, JSON data format, or any combination thereof.

[0283] State 27: The method according to any one of States 47 to 59, wherein the device includes a UE, a base station, a network entity, a relay device, a sidelink device, or an IAB node.

[0284] State 28: An apparatus for wireless communication at a device in a wireless network, comprising a processor and memory coupled to the processor; wherein the processor is configured to cause the apparatus to perform a method according to any one of states 34 to 46.

[0285] Format 29: An apparatus for wireless communication at a device in a wireless network, comprising at least one component for performing the method according to any one of Formats 34 to 46.

[0286] Format 30: A non-transitory computer-readable medium storing code for wireless communication at a device in a wireless network, the code including instructions executable by a processor to perform the method according to any one of Formats 34 to 46.

[0287] State 31: An apparatus for wireless communication at a network entity, comprising a processor and memory coupled to the processor, wherein the processor is configured to cause the apparatus to perform a method according to any one of states 47 to 60.

[0288] Format 32: An apparatus for wireless communication at a network entity, comprising at least one component for performing the method according to any one of Formats 47 to 60.

[0289] Format 33: A non-transitory computer-readable medium storing code for wireless communication at a network entity, the code including instructions executable by a processor to perform the method according to any one of Formats 47 to 60.

[0290] Sample 34: A method for wireless communication at a device in a wireless network, comprising the steps of: receiving a configuration message for the device, the configuration message indicating one or more neural network models for the device; receiving an instruction for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and, according to the operation sequence, performing the signal processing procedure for the at least one neural network model using signals received at the device.

[0291] State 35: The method according to State 34 also includes the step of: receiving a signal transmission that configures the device with an operation set, the operation set including one or more operations for the operation sequence of the at least one neural network model.

[0292] State 36: According to any one of states 34 to 35, receiving the instruction on the operation sequence includes: receiving the instruction on the operation sequence for the signal processing procedure in the configuration message, wherein the configuration message includes an operation set, the operation set including all operations of the operation sequence for the at least one neural network model.

[0293] State 37: According to any one of states 34 to 36, receiving the instruction on the operation sequence includes: receiving a second configuration message for the device, the second configuration message indicating: the at least one neural network model, the instruction on the operation sequence, and an operation set including all operations of the operation sequence for the at least one neural network model.

[0294] State 38: According to the method of any one of states 34 to 37, the receipt of the instruction on the operation sequence includes: receiving a set of input parameters, a set of output parameters, or both of the operation sequence for the at least one neural network model.

[0295] State 39: The method according to any one of states 34 to 38 also includes the step of: receiving an instruction for a mapping between the one or more neural network models and the set of operating conditions, wherein, according to the operation sequence, performing the signal processing procedure for the at least one neural network model using the signal received at the device is at least partially based on the mapping between the one or more neural network models and the set of operating conditions.

[0296] State 40: According to the method of State 39, wherein the set of operating conditions includes an SNR range, bandwidth range, signal scaling range, channel delay profile, signal peak range, or any combination thereof associated with the signal received at the device.

[0297] State 41: The method according to any one of states 34 to 40 also includes the step of: transmitting a message indicating the device's ability to support one or more operations for signal processing, wherein receiving the indication of the sequence of operations for the signal processing procedure for the at least one neural network model is at least partially based on the device's capability.

[0298] State 42: The method according to state 41, wherein the operation sequence includes operations supported by the device.

[0299] State 43: The method according to any one of states 41 to 42, wherein the message indicating that the device supports the capability for the one or more operations for signal processing includes: an indication of a threshold input dimension for each of the one or more operations for signal processing or a threshold running time for each of the one or more operations for signal processing.

[0300] State 44: According to any one of states 34 to 43, receiving the instruction on the operation sequence includes: receiving an RRC signal transmission or MAC-CE including the instruction on the operation sequence.

[0301] State 45: The method according to any one of states 34 to 44 also includes the step of: receiving an instruction for one or more data formats associated with one or more operations of the operation sequence, the one or more data formats including XML data format, JSON data format, or any combination thereof.

[0302] State 46: The method according to any one of states 34 to 45, wherein the device includes a UE, a base station, a network entity, a relay device, a sidelink device, or an IAB node.

[0303] Sample 47: A method for wireless communication at a base station, comprising the steps of: transmitting a configuration message to a device, the configuration message indicating one or more neural network models for the device; transmitting an instruction for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and transmitting a signal to the device at least in part based on the instruction for the operation sequence of the signal processing procedure for the at least one neural network model.

[0304] State 48: The method according to state 47 also includes the following steps: transmitting a signal transmission that configures the device with an operation set, the operation set including one or more operations for the operation sequence of the at least one neural network model.

[0305] State 49: The method according to any one of states 47 to 48 also includes the step of receiving a second operation sequence for executing a second signal transmission procedure at the base station.

[0306] State 50: According to any one of states 47 to 49, the transmission of the instruction on the operation sequence includes: transmitting an instruction on the operation sequence for the signal processing program in the configuration message, wherein the configuration message includes an operation set, the operation set including all operations of the operation sequence for the at least one neural network model.

[0307] State 51: According to any one of states 47 to 50, the transmission of the instruction to the operation sequence includes: transmitting a second configuration message to the device, the second configuration message indicating: the at least one neural network model, the instruction to the operation sequence, and an operation set including all operations for the operation sequence of the at least one neural network model.

[0308] State 52: According to the method of any one of states 47 to 51, the instruction of the transmission to the operation sequence includes: transmitting a set of input parameters, a set of output parameters, or both of the operation sequence for the at least one neural network model.

[0309] State 53: The method according to any one of states 47 to 52 also includes the step of: transmitting an indication of the mapping between the one or more neural network models and the set of operating conditions.

[0310] State 54: According to the method of State 53, the set of operating conditions includes the SNR range, bandwidth range, signal scaling range, channel delay profile, signal peak range, or any combination thereof associated with the signal transmitted to the device.

[0311] State 55: The method according to any one of states 47 to 54 also includes the step of: receiving a message indicating that the device supports the capability of one or more operations for signal processing, wherein the indication of receiving the sequence of operations for the signal processing procedure for the at least one neural network model is at least partially based on the capability of the device.

[0312] State 56: The method according to state 55, wherein the operation sequence includes operations supported by the device.

[0313] State 57: According to any one of states 55 to 56, the message indicating that the device supports the capability for the one or more operations for signal processing includes: an indication of a threshold input dimension for each of the one or more operations for signal processing or an indication of a threshold running time for each of the one or more operations for signal processing.

[0314] State 58: According to any one of states 47 to 57, the transmission of the indication to the operation sequence includes: transmitting an RRC signal transmission or MAC-CE including the indication to the operation sequence.

[0315] State 59: The method according to any one of states 47 to 58 also includes the step of: transmitting an indication of one or more data formats associated with one or more operations of the operation sequence, the one or more data formats including XML data format, JSON data format, or any combination thereof.

[0316] State 60: The method according to any one of states 47 to 59, wherein the device includes a UE, a base station, a network entity, a relay device, a sidelink device, or an IAB node.

[0317] State 61: An apparatus for wireless communication at a device in a wireless network, comprising: a processor; a memory coupled to the processor; and instructions stored in the memory and executable by the processor to cause the apparatus to perform the method according to any one of states 34 to 46.

[0318] Format 62: An apparatus for wireless communication at a device in a wireless network, comprising at least one component for performing the method according to any one of Formats 34 to 46.

[0319] Format 63: A non-transitory computer-readable medium storing code for wireless communication at a device in a wireless network, the code including instructions executable by a processor to perform the method according to any one of Formats 34 to 46.

[0320] Sample 64: An apparatus for wireless communication at a base station, comprising: a processor; a memory coupled to the processor; and instructions stored in the memory and executable by the processor to cause the apparatus to perform a method according to any one of Samples 47 to 60.

[0321] Sample 65: An apparatus for wireless communication at a base station, comprising at least one component for performing the method according to any one of Samples 47 to 60.

[0322] Sample 66: A non-transitory computer-readable medium storing code for wireless communication at a base station, the code including instructions executable by a processor to perform the method according to any one of Samples 47 to 60.

[0323] It should be noted that the methods described herein describe possible implementations, and the operations and steps can be rearranged or otherwise modified, and other implementations are possible. Furthermore, variants from two or more methods can be combined.

[0324] Although various forms 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 technologies described herein are applicable to areas outside of LTE, LTE-A, LTE-A Pro, or NR networks. For example, the described technologies 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.

[0325] The information and signals described herein can be represented using any of a variety of different techniques and methods. For example, data, instructions, commands, information, signals, bits, symbols and chips mentioned throughout the description may be represented by voltage, current, electromagnetic waves, magnetic fields or particles, light fields or particles, or any combination thereof.

[0326] The various illustrative blocks and elements described in connection with the disclosure herein may be implemented or executed using a general-purpose processor, DSP, ASIC, CPU, FPGA or other programmable logic device, individual gate or transistor logic, individual hardware element or any combination thereof designed to perform the functions described herein. The 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, a combination of one or more microprocessors and a DSP core, or any other such configuration).

[0327] The functions described herein can be implemented using hardware, processor-executed software, firmware, or any combination thereof. If implemented using processor-executed software, such functions can be stored as one or more instructions or codes on or transmitted via a computer-readable medium. Other examples and implementations are within the scope of this document and the appended claims. For example, due to the nature of software, the functions described herein can be implemented using processor-executed software, hardware, firmware, hardwiring, or any combination thereof. Features implementing the functions can also be located at various physical locations, including being distributed such that portions of the functions are implemented at different physical locations.

[0328] Computer-readable media includes both non-transitory computer storage media and communication media, with communication media including any media that facilitates the transfer of computer programs from one place to another. Non-transitory storage media can be any available media that can be accessed by a general-purpose computer or a special-purpose computer. By way of example, and not limitation, non-transitory computer-readable media can include RAM, ROM, electronically erasable programmable ROM (EEPROM), flash memory, compressed optical disc (CD) ROM or other optical disc storage, magnetic disk storage or other magnetic storage devices, or any other non-transitory media that can be used to carry or store desired program code components in the form of instructions or data structures, and that can be accessed by a general-purpose or special-purpose computer, or a general-purpose or special-purpose processor. Furthermore, any connection is properly 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 within the definition of computer-readable media. As used herein, magnetic disks and optical disks include CDs, laser discs, optical discs, digital versatile discs (DVDs), floppy disks, and Blu-ray discs, where magnetic disks typically copy data magnetically, while optical discs use lasers to optically copy data. The combination of the above is also included within the scope of computer-readable media.

[0329] As used herein (including in a request), the word "or" as used in a list of projects (e.g., a list of projects ending with a phrase such as "at least one of" or "one or more of") indicates an inclusive list such that a list of, for example, 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 closed set of conditions. For example, without departing from the scope of this document, an exemplary step described as "based on condition A" could be based on both condition A and condition B. In other words, as used herein, the phrase "based on" should be interpreted in the same way as the phrase "at least partially based on".

[0330] The term "determine" or "determining" encompasses a wide variety of actions, and therefore, "determining" can include calculation, operation, processing, deduction, research, examination (e.g., examining in a table, database, or other data structure), ascertainment, etc. Furthermore, "determining" can include receiving (e.g., receiving information), accessing (e.g., accessing data in memory), etc. Additionally, "determining" can include parsing, selecting, choosing, creating, and other similar actions.

[0331] In the accompanying drawings, similar elements or features may have the same element symbol. Furthermore, various elements of the same type can be distinguished by a dash and a second mark following the element symbol, which is used to differentiate between similar elements. If only a first element symbol is used in the specification, the description applies to any one of the similar elements having the same first element symbol, without regard to the second element symbol or other subsequent element symbols.

[0332] The exemplary configurations described herein, in conjunction with the accompanying drawings, are not intended to represent all instances that can be implemented or that fall within the scope of the requested items. As used herein, the term "instance" means "used as an example, instance, or illustration," not "preferred" or "superior to other instances." The detailed description includes specific details for the purpose of providing an understanding of the described techniques. However, these techniques may be implemented without such specific details. In some instances, known structures and devices are illustrated in block diagram form to avoid obscuring the concept of the described instances.

[0333] The description herein is provided to enable those skilled in the art to implement or use the content of this application. Various modifications to the content of this application will be readily apparent to those skilled in the art, and the overall principles defined herein can be applied to other variations without departing from the scope of the content of this application. Therefore, the content of this application is not limited to the examples and designs described herein, but is given the broadest scope consistent with the principles and novel features revealed herein. [Simplified Explanation of the Diagram]

[0070] Figures 1 and 2 illustrate examples of wireless communication systems that support one or more of the techniques for instructing signal processing procedures for deploying neural network models on a network, according to the content of this case.

[0071] Figure 3 illustrates an example of a flowchart of a technique for instructing a signal processing procedure for deploying a neural network model, according to one or more states of the present invention.

[0072] Figure 4 illustrates an example of a machine learning process that supports one or more states according to the content of this case and is used to indicate a signal processing procedure for deploying a neural network model on a network.

[0073] Figure 5 illustrates an example of a process flow that supports one or more states according to the content of this case for instructing a signal processing procedure for deploying a neural network model on a network.

[0074] Figures 6 and 7 illustrate block diagrams of devices that support one or more states of a technology for instructing signal processing procedures for deploying neural network models in accordance with the contents of this case.

[0075] Figure 8 is a block diagram illustrating a communication manager that supports one or more states according to the content of this case for instructing signal processing procedures for deploying neural network models on a network.

[0076] Figure 9 illustrates a system including a device for instructing a signal processing program for deploying a neural network model according to one or more of the contents of this case.

[0077] Figures 10 and 11 illustrate block diagrams of a device that supports one or more states of a technology for instructing a signal processing procedure for deploying a neural network model in accordance with the contents of this case.

[0078] Figure 12 is a block diagram illustrating a communication manager that supports one or more states according to the content of this case for instructing signal processing programs for deploying neural network models on a network.

[0079] Figure 13 illustrates a system including a device for instructing a signal processing program for deploying a neural network model according to one or more of the contents of this case.

[0080] Figures 14 to 19 illustrate flowcharts of a method for instructing a signal processing procedure for deploying a neural network model, according to one or more states of the present invention. [Biomaterial Storage]

[0335] Domestic storage information (please note in order of storage institution, date, and number): None. International storage information (please note in order of storage country, institution, date, and number): None.

Claims

1. An apparatus for wireless communication at a device in a wireless network, comprising: One processor; The processor and memory coupled to the processor are configured to: obtain a configuration message for the device, wherein the configuration message indicates one or more neural network models for the device; obtain an instruction for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and, according to the operation sequence, execute the signal processing procedure for the at least one neural network model using a signal obtained at the device.

2. The apparatus according to claim 1, wherein the processor is also configured to: receive a signal transmission configuring the apparatus to have an operation set, the operation set including one or more operations for the operation sequence for the at least one neural network model.

3. The apparatus according to claim 1, wherein, in order to obtain the instruction on the operation sequence, the processor is configured to: obtain the instruction on the operation sequence for the signal processing program in the configuration message, wherein the configuration message includes an operation set that includes all operations of the operation sequence for the at least one neural network model.

4. The apparatus according to claim 1, wherein, in order to obtain the instruction on the operation sequence, the processor is configured to: obtain a second configuration message for the apparatus, wherein the second configuration message indicates: the at least one neural network model, the instruction on the operation sequence, and an operation set including all operations for the operation sequence for the at least one neural network model.

5. The apparatus according to claim 1, wherein, in order to obtain the instruction on the operation sequence, the processor is configured to: obtain a set of input parameters, a set of output parameters, or both, for one or more operations of the operation sequence for the at least one neural network model.

6. The apparatus according to claim 1, wherein the processor is also configured to: obtain an indication of a mapping between the one or more neural network models and a set of operating conditions, wherein the signal processing procedure for the at least one neural network model is at least partially based on the mapping between the one or more neural network models and the set of operating conditions, and is executed using the signal obtained at the device according to the sequence of operations.

7. The apparatus according to claim 6, wherein the set of operating conditions includes a signal-to-noise ratio range, a bandwidth range, a signal scaling range, a channel delay profile, a signal peak range, or any combination thereof associated with the signal obtained at the apparatus.

8. The apparatus according to claim 1, wherein the processor is also configured to: output a message indicating that the device supports a capability for one or more operations of one or more signal processing programs, wherein the indication of the sequence of operations of the signal processing program for the at least one neural network model is obtained at least in part based on the capability of the device.

9. The apparatus according to claim 8, wherein the sequence of operations includes one or more operations supported by the apparatus.

10. The apparatus according to claim 8, wherein the message instructing the device to support the capability for the one or more operations of the one or more signal processing procedures includes: A threshold input dimension for each of the one or more operations of the one or more signal processing programs, or an indication of a threshold running time for each of the one or more operations of the one or more signal processing programs.

11. The apparatus according to claim 1, wherein, in order to obtain the instruction on the operation sequence, the processor is configured to: obtain a Radio Resource Control (RRC) signaling transmission or a Media Access Control (MAC) control element (MAC-CE) including the instruction on the operation sequence.

12. The apparatus according to claim 1 also includes: A line is configured to: obtain an indication of one or more data formats associated with one or more operations in the sequence of operations, wherein the one or more data formats include an Extensible Markup Language data format, a JavaScript Object Markup Language data format, or any combination thereof.

13. The apparatus according to claim 1, wherein the apparatus includes a user equipment (UE), a base station, a network entity, a relay device, a side link device, or an integrated access and backhaul (IAB) node.

14. An apparatus for wireless communication at a network entity, comprising: One processor; The processor and memory coupled to the processor are configured to: output a configuration message to a device, wherein the configuration message indicates one or more neural network models for the device; output an indication of an operation sequence for a signal processing procedure for at least one of the one or more neural network models, wherein the signal processing procedure includes one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and output a signal to the device based at least in part on the indication of the operation sequence of the signal processing procedure for the at least one neural network model.

15. The apparatus according to claim 14, wherein the processor is also configured to: output a signal transmission configuring the apparatus to have a set of operations, the set of operations including one or more operations for the sequence of operations for the at least one neural network model.

16. The apparatus according to claim 14, wherein the processor is also configured to: obtain a second sequence of operations for executing a second signaling procedure at the network entity.

17. The apparatus according to claim 14, wherein, in order to output the instruction for the operation sequence, the processor is configured to: output the instruction for the operation sequence for the signal processing program in a configuration message, wherein the configuration message includes an operation set that includes all operations of the operation sequence for the at least one neural network model.

18. The apparatus according to claim 14, wherein, in order to output the instruction for the operation sequence, the processor is configured to: output a second configuration message to the device, wherein the second configuration message indicates: the at least one neural network model, the instruction for the operation sequence, and an operation set including all operations for the operation sequence for the at least one neural network model.

19. The apparatus according to claim 14, wherein, in order to output the instruction on the sequence of operations, the processor is configured to: output a set of input parameters, a set of output parameters, or both, for one or more operations of the sequence of operations for the at least one neural network model.

20. The apparatus according to claim 14, wherein the processor is also configured to: output an indication of a mapping between the one or more neural network models and a set of operating conditions.

21. The apparatus according to claim 20, wherein the set of operating conditions includes a signal-to-noise ratio range, a bandwidth range, a signal scaling range, a channel delay profile, a signal peak range, or any combination thereof associated with the signal transmitted to the apparatus.

22. The apparatus according to claim 14, wherein the processor is also configured to: obtain a message indicating that the device supports a capability for one or more operations of one or more signal processing programs, wherein obtaining the indication of the sequence of operations of the signal processing program for the at least one neural network model is at least partially based on the capability of the device.

23. The apparatus according to claim 22, wherein the sequence of operations includes one or more operations supported by the apparatus.

24. The apparatus according to claim 22, wherein the message instructing the device to support the capability for the one or more operations of the one or more signal processing procedures includes: A threshold input dimension for each of the one or more operations of the one or more signal processing programs, or an indication of a threshold running time for each of the one or more operations of the one or more signal processing programs.

25. The apparatus according to claim 14, wherein, in order to output the indication of the operation sequence, the processor is configured to: output a Radio Resource Control (RRC) signal transmission or a Media Access Control (MAC) control element (MAC-CE) including the indication of the operation sequence.

26. The apparatus according to claim 14 also includes: A line is configured to output an indication of one or more data formats associated with one or more operations in the sequence of operations, wherein the one or more data formats include an Extensible Markup Language data format, a JavaScript Object Markup Language data format, or any combination thereof.

27. The apparatus according to claim 14, wherein the apparatus includes a user equipment (UE), a base station, a network entity, a relay device, a side link device, or an integrated access and backhaul (IAB) node.

28. A method for wireless communication at a device in a wireless network, comprising the steps of: obtaining a configuration message for the device, the configuration message indicating one or more neural network models for the device; obtaining an instruction for an operation sequence of a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and, according to the operation sequence, performing the signal processing procedure for the at least one neural network model using a signal obtained at the device.

29. The method according to claim 28 also includes the steps of: obtaining a signal transmission configuring the device with an operation set, the operation set including one or more operations for the operation sequence for the at least one neural network model.

30. A method for wireless communication at a network entity, comprising the steps of: outputting a configuration message to a device, the configuration message indicating one or more neural network models for the device; outputting an indication of an operation sequence for a signal processing procedure for at least one of the one or more neural network models, the signal processing procedure including one of an input preprocessing procedure or an output preprocessing procedure associated with the at least one neural network model; and outputting a signal to the device at least in part based on transmitting the indication of the operation sequence of the signal processing procedure for the at least one neural network model.

Images