Communication method and communication apparatus

Abstract

A communication method and a communication apparatus, which can improve the flexibility and efficiency of model alignment of two communication parties. The method comprises the following steps: a first communication apparatus sending first information, wherein the first information indicates a model alignment mode supported by the first communication apparatus; the first communication apparatus receiving second information, wherein the second information indicates a target model alignment mode, the target model alignment mode is a model alignment mode supported by a second communication apparatus, and the target model alignment mode is a model alignment mode among model alignment modes supported by the first communication apparatus; and the first communication apparatus performing model alignment on the basis of the target model alignment mode.

Description

Communication method and communication device

[0001] This application claims priority to the Chinese patent application filed with the State Intellectual Property Office on December 15, 2023, with application number 202311734447.7 and application name “Communication Method and Communication Device”, the entire contents of which are incorporated by reference into this application. Technical Field

[0002] The present application relates to the field of communication technology, and in particular to a communication method and a communication device. Background Art

[0003] Neural network-based transmitters and receivers, such as encoders and decoders, can be trained using real-world data to optimize transmission signal design and reception performance based on specific scenarios. Neural network transceivers can be used for physical layer signal processing, such as symbol modulation and demodulation, channel coding and decoding, and pilot-channel estimation; they can also be used for data processing, such as channel state information compression and reconstruction.

[0004] The transmitter and receiver neural networks can be jointly optimized to achieve optimal performance. For example, between communicating devices, the transmitter / receiver trained on one end can be adapted to the receiver / transmitter trained on the other end to achieve optimal end-to-end performance. However, differences in capabilities and vendors between the communicating parties can make joint optimization inefficient or even impossible. Summary of the Invention

[0005] The present application provides a communication method and a communication device to improve the flexibility and efficiency of model alignment between communicating parties.

[0006] A first aspect provides a communication method. The method is performed by a first communication device (the first communication device may be a terminal device or a network device), or the method is performed by some components (such as a processor, chip, or chip system) in the first communication device, or the method can also be implemented by a logic module or software that can implement all or part of the functions of the first communication device.

[0007] The method includes: a first communication device sends first information, and the first information indicates a model alignment mode supported by the first communication device. In other words, the first information indicates the model alignment capability of the first communication device. The model alignment mode includes at least one of the following: a model-based alignment mode, a gradient-based alignment mode, and a data set-based alignment mode. The first communication device receives second information, and the second information indicates a target model alignment mode. The target model alignment mode is a model alignment mode supported by the second communication device, and the target alignment mode is a model alignment mode among the model alignment modes supported by the first communication device. Then, the first communication device performs model alignment based on the target model alignment mode. The first communication device and the second communication device exchange the model alignment modes supported by themselves, and then determine the target model alignment model supported by both the first communication device and the second communication device for model alignment, so that the model alignment mode can be flexibly selected according to the alignment capability of the communication device, which can improve the efficiency and success rate of model alignment.

[0008] In one possible implementation, the first communication device performs model alignment based on a target model alignment mode, including: the first communication device sends or receives target information corresponding to the target model alignment mode, and the target information is used to train the first model. The target alignment mode is a model-based alignment mode, and the target information includes the second model. The target alignment mode is a gradient-based alignment mode, and the target information includes the gradient. The target alignment mode is a dataset-based alignment mode, and the target information includes the dataset. Under different model alignment modes, the data used to train the first model and the requirements for the communication device are different. The target model alignment mode is determined according to the capabilities of the first communication device and the second communication device, and training is performed based on the target model alignment mode, making model alignment more flexible.

[0009] In one possible implementation, the target alignment mode is a model-based alignment mode, and the target information includes a second model. The first model is the model of the first communication device, and the second model is the model of the second communication device; or, the first model is the model of the second communication device, and the second model is the model of the first communication device. When both communicating parties support the model-based alignment mode, that is, when one of the communicating parties can disclose its own model and the other party can parse and run the model sent by the other party, it can be determined that the target model alignment mode is the model-based alignment mode. The model-based alignment mode can perform offline model training, which can reduce the requirements on the computing power of the communication device, and the communicating parties do not need to frequently exchange information, which can reduce the air interface overhead in the model alignment process.

[0010] In one possible implementation, the target alignment mode is a gradient-based alignment mode, and the target information includes a gradient, which is obtained based on the output of the first model and the second model; the first model is a sending neural network model, and the second model is a receiving neural network model corresponding to the first model. The first model is a model of the first communication device, and the second model is a model of the second communication device; or, the first model is a model of the second communication device, and the second model is a model of the first communication device. When both communicating parties support the gradient-based alignment mode, that is, when one of the communicating parties can calculate the reverse gradient and the other party can parse the reverse gradient and update the model, the target model alignment mode can be determined to be a gradient-based alignment mode. The model-based alignment mode can perform online model training, which can improve the efficiency of model alignment. In addition, model alignment can be completed without the communicating parties disclosing their own models, thereby improving the flexibility of model alignment.

[0011] In one possible implementation, the target alignment mode is a dataset-based alignment mode, and the target information includes a dataset, which includes input data and / or output data of the second model. The first model is a model of the first communication device, and the second model is a model of the second communication device; or, the first model is a model of the second communication device, and the second model is a model of the first communication device. When both communicating parties support the dataset-based alignment mode, that is, when one of the communicating parties can provide a dataset for training and the other party can use the dataset to train a model, the target model alignment mode can be determined to be a dataset-based alignment mode. The dataset-based alignment mode allows for offline model training, which can reduce the capability requirements of the communication device. Furthermore, model alignment can be completed without the communicating parties disclosing their own models, thereby increasing the flexibility of model alignment.

[0012] In a possible implementation, performing model alignment based on a target model alignment mode further includes: the first communication device sending or receiving verification information, where the verification information is obtained based on the first model, and the verification information is used to verify whether the first model is aligned with the second model.

[0013] In one possible implementation, the method further includes: if the first model and the second model are aligned, the first communication device determining an association relationship between the first model and the second model. Thus, based on the association relationship between the first model and the second model, model selection, switching, activation, and monitoring in subsequent processes are facilitated, thereby facilitating application and management of the first model and the second model.

[0014] In one possible implementation, the first information further indicates the priority of the model alignment modes supported by the first communication device, with the target model alignment mode being the highest-priority model alignment mode supported by the first and second communication devices. Thus, when multiple candidate model alignment modes are available, the model alignment mode desired by the first communication device can be determined based on the priority, thereby improving the flexibility and efficiency of model alignment.

[0015] In one possible implementation, the first information includes at least one of the following: a first bitmap indicating the model alignment modes supported by the first communication device; an index of the model alignment modes supported by the first communication device. The second information includes at least one of the following: a second bitmap indicating the target model alignment mode; an index of the target model alignment mode. Thus, the first information can accurately indicate the model alignment modes supported by the first communication device. The second information can also accurately indicate the target model alignment mode determined by the second communication device, thereby improving the efficiency and success rate of model alignment.

[0016] A second aspect provides a communication method. The method is performed by a second communication device (which may be a terminal device or a network device), or by some components (such as a processor, chip, or chip system) in the second communication device, or by a logic module or software that can implement all or part of the functions of the second communication device.

[0017] The method includes: a second communication device receiving first information indicating a model alignment mode supported by the first communication device; the second communication device sending second information indicating a target model alignment mode, the target model alignment mode being a model alignment mode supported by the second communication device and a model alignment mode among the model alignment modes supported by the first communication device; and the second communication device performing model alignment based on the target model alignment mode.

[0018] In a possible implementation, the model alignment mode includes at least one of the following: a model-based alignment mode; a gradient-based alignment mode; and a dataset-based alignment mode.

[0019] In one possible implementation, the second communication device performs model alignment based on the target model alignment mode, including: the second communication device sends or receives target information corresponding to the target model alignment mode, and the target information is used to train the first model.

[0020] In one possible implementation, the target alignment mode is a model-based alignment mode, and the target information includes a second model; wherein the first model is a model of the first communication device, and the second model is a model of the second communication device; or, the first model is a model of the second communication device, and the second model is a model of the first communication device.

[0021] In one possible implementation, the target alignment mode is a gradient-based alignment mode, and the target information includes a gradient, which is obtained based on the output of the first model and the second model; the first model is a sending neural network model, and the second model is a receiving neural network model corresponding to the first model; wherein the first model is a model of the first communication device, and the second model is a model of the second communication device; or, the first model is a model of the second communication device, and the second model is a model of the first communication device.

[0022] In one possible implementation, the target alignment mode is a data set-based alignment mode, the target information includes a data set, and the data set includes input data and / or output data of the second model; wherein the first model is a model of the first communication device, and the second model is a model of the second communication device; or, the first model is a model of the second communication device, and the second model is a model of the first communication device.

[0023] In one possible implementation, the second communication device performs model alignment based on the target model alignment mode, further comprising: the second communication device sending or receiving verification information, the verification information being obtained based on the first model, and the verification information being used to verify whether the first model is aligned with the second model.

[0024] In one possible implementation, if the first model is aligned with the second model, the second communication device associates the first model with the second model.

[0025] In a possible implementation, the first information further indicates the priority of the model alignment mode supported by the first communication device, and the target model alignment mode is the model alignment mode with the highest priority among the model alignment modes supported by the first communication device and the second communication device.

[0026] In one possible implementation, the first information includes at least one of the following: a first bitmap indicating a model alignment mode supported by the first communication device; an index of the model alignment mode supported by the first communication device. The second information includes at least one of the following: a second bitmap indicating a target model alignment mode; an index of the target model alignment mode.

[0027] A third aspect provides a communication device. The communication device has the functionality to implement the behavior described in the method example of the first aspect. The beneficial effects can be found in the description of the first aspect and are not further described here. The communication device may be the first communication device described in the first aspect, or it may be a device capable of supporting the first communication device described in the first aspect to implement the functionality required by the method provided in the first aspect, such as a chip or chip system.

[0028] In one possible design, the communication device includes corresponding means or modules for performing the method of the first aspect. For example, the communication device includes a processing unit (sometimes also referred to as a processing module) and / or a transceiver unit (sometimes also referred to as a transceiver module). These units (modules) can perform the corresponding functions in the above-mentioned method example of the first aspect. For details, please refer to the detailed description in the method example, which is not repeated here.

[0029] In a fourth aspect, an embodiment of the present application provides a communication device having the function of implementing the behavior in the method example of the second aspect above. The beneficial effects can be found in the description of the second aspect and are not repeated here. The communication device may be the second communication device in the second aspect, or the communication device may be a device capable of supporting the second communication device in the second aspect to implement the functions required by the method provided in the second aspect, such as a chip or chip system.

[0030] In one possible design, the communication device includes corresponding means or modules for performing the method of the second aspect. For example, the communication device includes a processing unit (sometimes also referred to as a processing module) and / or a transceiver unit (sometimes also referred to as a transceiver module). These units (modules) can perform the corresponding functions in the above-mentioned method example of the second aspect. For details, please refer to the detailed description in the method example, which is not repeated here.

[0031] In a fifth aspect, an embodiment of the present application provides a communication device, which may be the communication device in the third or fourth aspect of the above-mentioned embodiment, or a chip or chip system provided in the communication device in the third or fourth aspect. The communication device includes a communication interface and a processor, and optionally, further includes a memory. The memory is used to store computer programs, instructions, or data, and the processor is coupled to the memory and the communication interface. When the processor reads the computer program, instructions, or data, the communication device executes the method performed by the terminal device or network device in the above-mentioned method embodiment.

[0032] In a sixth aspect, an embodiment of the present application provides a communication device, comprising at least one processor and, optionally, a memory, wherein the at least one processor is coupled to the memory. The at least one processor is configured to execute the method described in the first aspect or the second aspect.

[0033] In a seventh aspect, an embodiment of the present application provides a chip system, which includes a processor and may also include a memory and / or a communication interface, for implementing the method described in the first aspect or the second aspect. In one possible implementation, the chip system also includes a memory for storing program instructions and / or data. The chip system can be composed of a chip, or it can include a chip and other discrete devices.

[0034] In an eighth aspect, an embodiment of the present application provides a communication system, comprising a communication device for executing the method described in the first aspect and a communication device for executing the method described in the second aspect. The communication device for executing the method described in the first aspect is, for example, the first communication device described in the first aspect, and the communication device for executing the method described in the second aspect is, for example, the second communication device described in the second aspect.

[0035] In a ninth aspect, the present application provides a computer-readable storage medium storing a computer program. When the computer program is executed, the method of any one of the first to second aspects described above is implemented.

[0036] In a tenth aspect, a computer program product is provided, comprising: a computer program code, wherein when the computer program code is run, the method in any one of the first to second aspects is executed.

[0037] Among them, the technical effects brought about by any design method in the second to tenth aspects can refer to the technical effects brought about by the different design methods in the above-mentioned first aspect, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] FIG1a is a schematic diagram of a communication system provided by the present application;

[0039] FIG1b is a schematic diagram of another communication system provided by the present application;

[0040] FIG2 is a flow chart of a communication method provided by the present application;

[0041] FIG3a is a flow chart of a method for aligning an interactive model provided by the present application;

[0042] FIG3 b is a flow chart of another method for interaction model alignment mode provided by the present application;

[0043] FIG4 is a flow chart of another communication method provided by the present application;

[0044] FIG5 is a flow chart of another communication method provided by the present application;

[0045] FIG6 is a flow chart of another communication method provided by the present application;

[0046] FIG7 is a flow chart of another communication method provided by the present application;

[0047] FIG8 is a schematic diagram of a communication device provided by the present application;

[0048] FIG9 is another schematic diagram of a communication device provided by the present application;

[0049] FIG10 is another schematic diagram of the communication device provided by the present application;

[0050] FIG11 is another schematic diagram of a communication device provided by the present application;

[0051] FIG12 is another schematic diagram of the communication device provided in this application. DETAILED DESCRIPTION

[0052] First, some of the terms used in the embodiments of the present application are explained to facilitate understanding by those skilled in the art.

[0053] (1) Terminal device: It can be a wireless terminal device that can receive network device scheduling and instruction information. The wireless terminal device can be a device that provides voice and / or data connectivity to the user, or a handheld device with wireless connection function, or other processing device connected to a wireless modem.

[0054] Terminal devices can communicate with one or more core networks or the Internet via a radio access network (RAN). Terminal devices can be mobile terminal devices, such as mobile phones (also known as "cellular" phones, mobile phones), computers, and data cards. For example, they can be portable, pocket-sized, handheld, computer-built-in, or vehicle-mounted mobile devices that exchange voice and / or data with the radio access network. Examples include personal communication service (PCS) phones, cordless phones, Session Initiation Protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (PDAs), tablet computers, and computers with wireless transceiver capabilities. Wireless terminal equipment can also be called system, subscriber unit, subscriber station, mobile station, mobile station (MS), remote station, access point (AP), remote terminal equipment (remote terminal), access terminal equipment (access terminal), user terminal equipment (user terminal), user agent, subscriber station (SS), customer premises equipment (CPE), terminal, user equipment (UE), mobile terminal (MT), etc.

[0055] As an example and not a limitation, in the embodiments of the present application, the terminal device may also be a wearable device. Wearable devices may also be referred to as wearable smart devices or smart wearable devices, etc., which are a general term for wearable devices that are intelligently designed and developed using wearable technology for daily wear, such as glasses, gloves, watches, clothing, and shoes. A wearable device is a portable device that is worn directly on the body or integrated into the user's clothes or accessories. Wearable devices are not only hardware devices, but also achieve powerful functions through software support, data interaction, and cloud interaction. Broadly speaking, wearable smart devices include those that are fully functional, large in size, and can achieve complete or partial functions without relying on smartphones, such as smart watches or smart glasses, etc., as well as those that only focus on a certain type of application function and need to be used in conjunction with other devices such as smartphones, such as various smart bracelets, smart helmets, and smart jewelry for vital sign monitoring.

[0056] The terminal may also be a drone, a robot, a terminal in device-to-device (D2D) communication, a terminal in vehicle-to-everything (V2X), a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal in industrial control, a wireless terminal in self-driving, a wireless terminal in remote medical, a wireless terminal in smart grid, a wireless terminal in transportation safety, a wireless terminal in smart city, a wireless terminal in smart home, etc.

[0057] In addition, the terminal device may also be a terminal device in a communication system that has evolved after the fifth generation (5G) communication system (e.g., a sixth generation (6G) communication system) or a terminal device in a future public land mobile network (PLMN). For example, the 6G network can further expand the form and function of 5G communication terminals. 6G terminals include but are not limited to vehicles, cellular network terminals (with integrated satellite terminal functions), drones, and Internet of Things (IoT) devices.

[0058] In an embodiment of the present application, the above-mentioned terminal device may also be a device having an AI model, which can process the data to be sent or the received signal based on the AI ​​model.

[0059] (2) Network equipment: It can be a device in a wireless network. For example, the network equipment can be a RAN node (or device) that connects a terminal device to a wireless network, which can also be called a base station. Currently, some examples of RAN equipment include: base station, evolved NodeB (eNodeB), gNB (gNodeB) in a 5G communication system, transmission reception point (TRP), evolved Node B (eNB), radio network controller (RNC), Node B (NB), home base station (e.g., home evolved Node B, or home Node B, HNB), base band unit (BBU), or wireless fidelity (Wi-Fi) access point AP, etc. In addition, in a network structure, the network equipment can include a centralized unit (CU) node, a distributed unit (DU) node, or a RAN device including a CU node and a DU node.

[0060] Alternatively, a RAN node can be a macro base station, micro base station, indoor base station, relay node, donor node, or a wireless controller in a cloud radio access network (CRAN) scenario. A RAN node can also be a server, wearable device, vehicle, or vehicle-mounted device. For example, the access network device in vehicle-to-everything (V2X) technology can be a roadside unit (RSU).

[0061] In another possible scenario, multiple RAN nodes collaborate to assist the terminal in achieving wireless access, and different RAN nodes respectively implement part of the functions of the base station. For example, the RAN node can be a centralized unit (CU), a distributed unit (DU), a CU-control plane (CP), a CU-user plane (UP), or a radio unit (RU). The CU and DU can be set separately, or they can be included in the same network element, such as a baseband unit (BBU). The RU can be included in a radio frequency device or radio frequency unit, such as a remote radio unit (RRU), an active antenna unit (AAU), or a remote radio head (RRH).

[0062] In different systems, CU (or CU-CP and CU-UP), DU or RU may also have different names, but those skilled in the art can understand their meanings. For example, in an open access network (open RAN, O-RAN or ORAN) system, CU may also be called O-CU (open CU), DU may also be called O-DU, CU-CP may also be called O-CU-CP, CU-UP may also be called O-CU-UP, and RU may also be called O-RU. For the convenience of description, this application takes CU, CU-CP, CU-UP, DU and RU as examples for description. Any unit of CU (or CU-CP, CU-UP), DU and RU in this application can be implemented by a software module, a hardware module, or a combination of a software module and a hardware module.

[0063] The communication between the access network device and the terminal device follows a certain protocol layer structure. The protocol layer may include a control plane protocol layer and a user plane protocol layer. The control plane protocol layer may include at least one of the following: a radio resource control (RRC) layer, a packet data convergence protocol (PDCP) layer, a radio link control (RLC) layer, a media access control (MAC) layer, or a physical (PHY) layer. The user plane protocol layer may include at least one of the following: a service data adaptation protocol (SDAP) layer, a PDCP layer, an RLC layer, a MAC layer, or a physical layer.

[0064] For the correspondence between network elements in the ORAN system and their achievable protocol layer functions, please refer to Table 1 below.

[0065] Table 1

[0066] The network device may be any other device that provides wireless communication functionality to the terminal device. The embodiments of this application do not limit the specific technology and device form used by the network device. For ease of description, the embodiments of this application do not limit this.

[0067] The network equipment may also include core network equipment, which may include, for example, a mobility management entity (MME), a home subscriber server (HSS), a serving gateway (S-GW), a policy and charging rules function (PCRF), and a public data network gateway (PDN gateway, P-GW) in a fourth generation (4G) network; and network elements such as an access and mobility management function (AMF), a user plane function (UPF), or a session management function (SMF) in a 5G network. In addition, the core network equipment may also include other core network equipment in a 5G network and a next generation network of a 5G network.

[0068] In an embodiment of the present application, the above-mentioned network device may also have a network node of an AI model, which can process data to be sent or received signals based on the AI ​​model.

[0069] In the embodiments of the present application, the apparatus for implementing the function of the network device may be a network device, or may be a device capable of supporting the network device in implementing the function, such as a chip system, which may be installed in the network device. In the technical solutions provided in the embodiments of the present application, the technical solutions provided in the embodiments of the present application are described by taking the apparatus for implementing the function of the network device as an example.

[0070] (3) AI model, also known as AI algorithm (or AI operator), is a general term for mathematical algorithms built based on the principles of artificial intelligence. It is also the basis for using AI to solve specific problems. Depending on the specific methods and / or technologies used to implement artificial intelligence, AI models can also be called machine learning models, deep learning models, or reinforcement learning models. Machine learning is a method of implementing artificial intelligence. The goal of this method is to design and analyze algorithms (also known as models) that allow computers to automatically "learn". The designed algorithms are called machine learning models. Machine learning models are a type of algorithm that automatically analyzes data to obtain patterns and uses these patterns to predict unknown data.

[0071] Currently, the typical structure of a deep learning model is a deep neural network. A neural network is a mathematical or computational model that mimics the structure and function of biological neural networks (the central nervous system of animals, particularly the brain). Neural networks perform computations by connecting a large number of neurons. A neural network can include multiple neural network layers with different functions, each with parameters and calculation rules. Different layers in a neural network have different names depending on the calculation formula or function. For example, a layer that performs convolution calculations is called a convolution layer, which is often used to extract features from input signals. A neural network can also be composed of multiple sub-neural networks. Different neural network structures can be applied to different scenarios (such as classification and recognition) or provide different results when used in the same scenario. Different neural network structures can specifically include one or more of the following: different number of layers in the neural network, different order of layers, and different weights, parameters, or calculation formulas in each layer. A variety of different neural networks with high accuracy are already available in the industry for application scenarios such as recognition and classification. Some neural networks can be trained with specific datasets and then used alone to complete a task or combined with other neural networks (or other functional modules) to complete a task.

[0072] (4) The terms "system" and "network" in the embodiments of the present application can be used interchangeably. "Multiple" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, "at least one of A, B and C" includes A, B, C, AB, AC, BC or ABC. In addition, unless otherwise specified, the ordinal numbers such as "first" and "second" mentioned in the embodiments of the present application are used to distinguish multiple objects, and are not used to limit the order, timing, priority or importance of multiple objects.

[0073] (5) “Sending” and “receiving” in the embodiments of the present application indicate the direction of signal transmission. For example, “sending information to XX” can be understood as the destination of the information being XX, which can include direct sending through the air interface, as well as indirect sending through the air interface by other units or modules. “Receiving information from YY” can be understood as the source of the information being YY, which can include direct receiving from YY through the air interface, as well as indirect receiving from YY through the air interface from other units or modules. “Sending” can also be understood as the “output” of the chip interface, and “receiving” can also be understood as the “input” of the chip interface.

[0074] In other words, sending and receiving can be performed between devices, for example, between a network device and a terminal device, or can be performed within a device, for example, sending or receiving between components, modules, chips, software modules or hardware modules within the device through a bus, wiring or interface.

[0075] It is understandable that information may be processed between the source and destination of information transmission, such as coding, modulation, etc., but the destination can understand the valid information from the source. Similar expressions in this application can be understood similarly and will not be repeated.

[0076] (6) In the embodiments of the present application, "indication" may include direct indication and indirect indication, and may also include explicit indication and implicit indication. The information indicated by a certain information (such as the indication information described below) is called information to be indicated. In the specific implementation process, there are many ways to indicate the information to be indicated, such as but not limited to, directly indicating the information to be indicated, such as the information to be indicated itself or the index of the information to be indicated. The information to be indicated may also be indirectly indicated by indicating other information, wherein the other information is associated with the information to be indicated; or only a part of the information to be indicated may be indicated, while the other part of the information to be indicated is known or agreed in advance. For example, the indication of specific information may be achieved by means of the arrangement order of each information agreed in advance (such as predefined by the protocol), thereby reducing the indication overhead to a certain extent. The present application does not limit the specific method of indication. It is understandable that for the sender of the indication information, the indication information can be used to indicate the information to be indicated, and for the receiver of the indication information, the indication information can be used to determine the information to be indicated.

[0077] In this application, unless otherwise specified, the same or similar parts between the various embodiments can refer to each other. In the various embodiments of this application, and the various methods / designs / implementations in each embodiment, if there is no special explanation and logical conflict, the terms and / or descriptions between different embodiments and the various methods / designs / implementations in each embodiment are consistent and can be referenced to each other. The technical features in different embodiments and the various methods / designs / implementations in each embodiment can be combined to form new embodiments, methods, or implementations according to their inherent logical relationships. The following description of the implementation methods of this application does not constitute a limitation on the scope of protection of this application.

[0078] The present application can be applied to a long term evolution (LTE) system, a new radio (NR) system, or a communication system evolved after 5G (such as 6G, etc.). The communication system includes at least one network device and / or at least one terminal device.

[0079] Please refer to Figure 1a, which is a schematic diagram of a communication system provided by the present application. Figure 1a exemplarily shows a network device and two terminal devices. Figure 1a takes the communication system as a cellular communication system and the network device as a base station as an example. The network device can communicate with each terminal device via wireless signals. The network device can send downlink signals to the terminal device, and the terminal device can send uplink signals to the network device. The terminal devices can also send and receive wireless signals via sidelinks (SL).

[0080] Please refer to Figure 1b, which is a schematic diagram of another communication system provided by the present application. Figure 1b shows an example of a network device and a terminal device. Figure 1b takes the communication system as a wireless local area network (WLAN) system and the network device as a wireless access point (AP) as an example. The network device can communicate with each terminal device via wireless signals. The network device can send downlink signals to the terminal device, and the terminal device can send uplink signals to the network device.

[0081] In a wireless communication system (such as the communication system shown in Figure 1a or Figure 1b), the communication node acting as the signal sender can subject the original data to be sent to multiple processing processes, including channel coding, modulation, etc.; correspondingly, the communication node acting as the signal receiver can subject the received signal to other processing processes corresponding to the multiple processing processes, including channel decoding, demodulation, etc., to restore the original data (or obtain an estimate of the original data). These processing processes can improve the reliability of data transmission.

[0082] After more than half a century of development, artificial intelligence (AI) technology has now fully entered the industrialization stage. AI technology has penetrated into various fields and industries, including wireless communications. Specifically, in wireless communication systems, AI models can be incorporated into the transmitter and receiver of devices (network devices or terminal devices). For example, the transmitter can use AI models to perform channel coding and modulation on the data to be transmitted, and the receiver can use AI models to perform channel decoding and demodulation on the received signal, thereby improving system flexibility, spectrum efficiency, and system stability.

[0083] To improve end-to-end performance between communicating parties, the AI ​​models in the transmitter and receiver can be jointly optimized. This means that the transmitter / receiver trained on one side is adapted to the receiver / transmitter trained on the other side. However, the devices in a communication system often come from different manufacturers, who may not necessarily be willing to make their trained models public. Furthermore, the capabilities of different devices vary. These factors can lead to low efficiency or even inability to perform joint optimization between the two communicating parties.

[0084] In view of this, the present application provides the following embodiments to enable flexible model alignment between devices and improve the model optimization efficiency of both communicating parties.

[0085] In one embodiment of the present application, a communication system includes a first communication device and a second communication device. The first communication device transmits its model alignment capabilities to the second communication device. The second communication device determines a target model alignment mode for both parties to align their models based on the alignment capabilities of the first communication device and its own alignment capabilities. Information is then exchanged based on the target model alignment mode to align the model in the first communication device with the model in the second communication device.

[0086] In this application, model alignment may also be referred to as model adaptation. Model alignment refers to training and adjusting the models of one or both communicating parties based on an end-to-end loss function so that the models of both communicating parties meet the end-to-end performance requirements. Alternatively, model alignment may also refer to using the model of one party as a reference to train the model of the other party so that the performance of the two is close. The model in this application refers to an AI model, which may also be referred to as a neural network model, a machine learning model, a deep learning model, or a reinforcement learning model. Specifically, the model may be, for example, a convolutional neural network (CNN), a recurrent neural network (RNN), a recursive neural network (RNN), a transformer, and the like.

[0087] As shown in Figure 2, Figure 2 is a flow chart of a communication method provided by this application. In this embodiment, the first communication device can be a network device and the second communication device is a terminal device, or the first communication device can be a terminal device and the second communication device is a network device, or the first communication device and the second communication device are both terminal devices. This embodiment includes the following steps:

[0088] S201: A first communication device sends first information indicating a model alignment mode supported by the first communication device. Correspondingly, a second communication device receives the first information.

[0089] The model alignment mode includes, for example, at least one of a model-based alignment mode, a gradient-based alignment mode, and a dataset-based alignment mode. The first information indicates the model alignment mode supported by the first communication device, that is, the first information may indicate the alignment capability of the first communication device. Accordingly, the model alignment mode supported by the first communication device may include at least one of a model-based alignment mode, a gradient-based alignment mode, and a dataset-based alignment mode.

[0090] Based on the model alignment mode, the communication device may send its own model to other communication devices, or the communication device may receive models from other communication devices and perform model alignment based on the received models.

[0091] Based on the alignment pattern of the gradients, the communication device can calculate the reverse gradient, or can decode the reverse gradient from other communication devices and train a model based on the gradient.

[0092] Based on the alignment mode of the data set, the communication device can provide sample data to the model of other communication devices for training, or can use sample data from other communication devices for model training to align the models of the two communicating parties.

[0093] The first information can be carried in a radio resource control (RRC) message, sidelink control information (SCI), a medium access control control element (MAC CE), downlink control information (DCI), or uplink control information (UCI), etc.

[0094] In one possible implementation, the first information includes, for example, a first bitmap, where a bit in the first bitmap may correspond to a model alignment mode, and the value in the bitmap indicates whether the first communication device supports the corresponding model alignment mode. For example, when the value in the bitmap is 1, it indicates that the first communication device supports the corresponding model alignment mode, and when the value in the bitmap is 0, it indicates that the first communication device does not support the corresponding model alignment mode. Alternatively, when the value in the bitmap is 0, it indicates that the first communication device supports the corresponding model alignment mode, and when the value in the bitmap is 1, it indicates that the first communication device does not support the corresponding model alignment mode. For example, taking the case where the first bitmap is 3 bits and the bit value is 1, indicating that the first communication device supports the corresponding model alignment mode, the first bit in the first bitmap corresponds to the model-based alignment mode, the second bit corresponds to the gradient-based alignment mode, and the third bit corresponds to the dataset-based alignment mode, and the value of the first bitmap is 101, it indicates that the first communication device supports the model-based alignment mode and the dataset-based alignment mode, but does not support the gradient-based alignment mode.

[0095] In another possible implementation, each model alignment mode has a corresponding index, and the first information includes, for example, the index of the supported model alignment mode. For example, the index of the model-based alignment mode is 01, the index of the gradient-based alignment mode is 10, and the index of the dataset-based alignment mode is 11. The inclusion of indexes 10 and 11 in the first information indicates that the first communication device supports the gradient-based alignment mode and the dataset-based alignment mode. It will be understood that the index of the model alignment mode here is only an example and should not be understood as a limitation of the present application.

[0096] In another possible implementation, the size of the first information may be 1 bit. For example, when the value of the first information is 1, it indicates that the first communication device supports all model alignment modes, and when the value of the first information is 0, it indicates that the first communication device supports some model alignment modes. Alternatively, for example, when the value of the first information is 0, it indicates that the first communication device supports all model alignment modes, and when the value of the first information is 1, it indicates that the first communication device supports some model alignment modes. This can reduce the signaling overhead of the first information.

[0097] Optionally, the first information also includes priority information corresponding to the model alignment mode supported by the first communication device, to instruct the second communication device to determine the model alignment mode with the highest priority as the target model alignment mode when there are multiple candidate model alignment modes (model alignment modes supported by both the first communication device and the second communication device).

[0098] Optionally, when the first communication device supports model-based alignment mode, the first information may also include, for example, a model description format of the model. The model description format may include, for example, a data structure and descriptions of tensors, network layers, and connection relationships. Only when the first and second communication devices can process the same model description format can the models be accurately parsed and aligned.

[0099] Optionally, when the first communication device supports a gradient-based alignment mode, the first information, for example, also includes the computing power of the first communication device. The computing power is, for example, the computing resources of the first communication device, which may be the currently available computing resources of the first communication device, or the total computing resources of the first communication device, or the computing resources currently used by the first communication device. The computing resources include at least one of XPU (for example, central processing unit (CPU), graphics processing unit (GPU), neural processing unit (NPU) or tensor processing unit (TPU)), memory resources or bandwidth resources. Since the gradient-based alignment mode performs online model alignment between the first communication device and the second communication device, the computing power of the communication device is relatively high. Therefore, the first information includes the computing power of the first communication device, which can assist the second communication device in determining whether the first communication device has the ability to perform model alignment based on gradient.

[0100] S202: The second communication device sends second information to the first communication device, where the second information indicates a target model alignment mode. Correspondingly, the first communication device receives the second information.

[0101] The second communication device determines a target model alignment mode based on the first information and the capabilities of the second communication device, and sends second information indicating the target model alignment mode to the first communication device. The target model alignment mode is a model alignment mode supported by the second communication device, and the target alignment mode is a model alignment mode among the model alignment modes supported by the first communication device. In other words, the target model alignment mode is an alignment mode supported by both the first communication device and the second communication device.

[0102] In one possible implementation, the second information may include a second bitmap. A bit in the second bitmap may correspond to a model alignment mode, and the value in the bit indicates whether the first communication device supports the corresponding model alignment mode. For example, a bit value of 0 indicates that the corresponding model alignment mode is target model alignment. Alternatively, a bit value of 1 indicates that the corresponding model alignment mode is target alignment.

[0103] In another possible implementation, each model alignment mode has a corresponding index, and the first information includes, for example, the index of the target model alignment mode.

[0104] Optionally, when both the first communication device and the second communication device support multiple model alignment modes, that is, multiple candidate model alignment modes, if the first information includes priority information of the model alignment mode, the second communication device determines that the model alignment mode with the highest priority among the candidate model alignment modes is the target model alignment mode.

[0105] Optionally, when both the first communication device and the second communication device support the model-based alignment mode, and the first information includes a model description format of the model, if the model description format is a model description format supported by the second communication device, the second communication device may determine the model-based alignment mode as the target model alignment mode, or determine the model-based alignment mode as the candidate model alignment mode. If the model description format is not a model description format supported by the second communication device, the model-based alignment mode is not used as the target model alignment mode.

[0106] Optionally, when both the first communication device and the second communication device support the gradient-based alignment mode, and the first information includes the computing capability of the first communication device, the second communication device can determine whether the computing capability of the first communication device supports aligning the model in the gradient-based alignment mode based on the complexity of the model to be aligned (for example, the number of parameters in the model, the quantization accuracy, etc.). If the computing capability of the first communication device supports aligning the model in the gradient-based alignment mode, the second communication device can determine the gradient-based alignment mode as the target model alignment mode, or determine the gradient-based alignment mode as the candidate model alignment mode. If the computing capability of the first communication device does not support aligning the model in the gradient-based alignment mode, the gradient-based alignment mode is not used as the target model alignment mode.

[0107] Optionally, the second information may further include parameter configurations for model training. For example, it may include at least one of quantization accuracy, dataset size, training rounds, and an information exchange period. The information exchange period may include, for example, the period for feedback gradients when the target model alignment mode is gradient-based model alignment mode.

[0108] S203: The first communication device and the second communication device perform model alignment based on the target model alignment mode.

[0109] The first communication device and the second communication device exchange target information corresponding to the target model alignment mode based on the target model alignment mode, and train the first model based on the target information to achieve alignment of the first model and the second model.

[0110] In one possible implementation, the first model is a model of a first communication device, and the second model is a model of a second communication device. That is, the first communication device trains the first model, and the second communication device provides target information corresponding to a target model alignment mode. This target information is used to train the first model to align the first model with the second model of the second communication device. In this case, as shown in FIG3a , the first information may be used to request the second communication device to perform model alignment. The first information is, for example, information in a model alignment request, or the first information is a model alignment request. Accordingly, the second information is information in a model alignment response corresponding to the model alignment request, or the second information is a model alignment response.

[0111] In another possible implementation, the first model is the model of the second communication device, and the second model is the model of the first communication device. That is, the second communication device trains the first model, and the first communication device provides target information corresponding to the target model alignment mode, and the target information is used for training the first model to align the second model with the first model of the first communication device. In this case, as shown in Figure 3b, the first information is used to send the model alignment capability of the first communication device. The first information is carried in a broadcast message, for example, or the first information can also be carried in a unicast message. The second information is used to request the first communication device to perform model alignment. The second information is, for example, the information in the model alignment request, or the second information is a model alignment request. Optionally, after receiving the second information, the first communication device can also send a model alignment response to the second communication device to confirm that the model alignment is performed based on the target model alignment mode.

[0112] The first model can be a sending model, and the second model can be a receiving model. Alternatively, the first model can be a receiving model, and the second model can be a sending model. Alternatively, both the first model and the second model can be sending models. Alternatively, both the first model and the second model can be receiving models.

[0113] The transmission model refers to the AI ​​model used in a transmitter for information compression, channel coding, or modulation. The reception model refers to the AI ​​model used in a receiver for information decompression, channel decoding, or demodulation. Since both the first communication device and the second notification device include a transmitter and a receiver, the first model can be a model in either the first communication device or the second communication device, and can be either a transmission model or a reception model.

[0114] The first model and the second model are related models. The second model is used to inversely process the data (signal) output by the first model, or the first model is used to inversely process the data (signal) output by the second model. Or the first model and the second model are models for implementing the same function. For example, when the first model is an AI model for information compression, the second model may be an AI model for information compression, or the second model may be an AI model for information decompression. For another example, when the first model is an AI model for channel coding, the second model may be an AI model for channel decoding, or the second model may be an AI model for channel decoding. For another example, the first model is an AI model for demodulation, the second model may be an AI model for modulation, or the second model may be an AI model for demodulation, and so on. Examples are not given one by one here.

[0115] In different model alignment modes, and whether the first model is a sending model or a receiving model, the process of performing model alignment between the first communication device and the second communication device is different.

[0116] In scenario 1, the first model is the sending model, and the target model alignment mode is the model-based alignment mode.

[0117] In the second scenario, the first model is the sending model, and the target model alignment mode is the dataset-based alignment mode.

[0118] In case three, the first model is the sending model, and the target model alignment mode is the gradient-based alignment mode.

[0119] In case 4, the first model is the receiving model, and the target model alignment mode is the model-based alignment mode.

[0120] In case five, the first model is the receiving model, and the target model alignment mode is the dataset-based alignment mode.

[0121] When the first model is the receiving model, the target model alignment mode does not include gradient-based alignment. This is because the end-to-end performance verification of the model is implemented on the receiving model side, that is, the loss function and gradients are implemented on the receiving model side, so there is no need for the sending model to provide gradients.

[0122] The following describes the process of performing model alignment between the first communication device and the second communication device in the above-mentioned scenarios 1 to 5 respectively.

[0123] For the above-mentioned scenario 1, as shown in Figure 4, Figure 4 is a flow chart of another communication method provided by the present application. This embodiment is implemented by communication device A and communication device B. The first model is the model in communication device A, and the second model is the model in communication device B. It should be noted that communication device A can be the first communication device in Figure 2, and communication device B is the second communication device in Figure 2. Alternatively, communication device A can be the second communication device in Figure 2, and communication device B is the first communication device in Figure 2. This embodiment includes the following steps:

[0124] S401: Communication device A sends a model alignment request to communication device B. Correspondingly, communication device B receives the model alignment request.

[0125] S402: Communication device B sends a model alignment response to communication device A. Correspondingly, communication device A receives the model alignment response.

[0126] In one possible implementation, communication device A and communication device B may determine that the target model alignment mode is a model-based alignment mode through the interaction process shown in FIG3a. For example, the model alignment request sent by communication device A to communication device B includes first information indicating the model alignment mode supported by communication device A. The model alignment response sent by communication device B to communication device A includes second information indicating the target model alignment mode determined by communication device B, which is the model-based alignment mode.

[0127] In another possible implementation, communication device A and communication device B may determine that the target model alignment mode is a model-based alignment mode through the interaction process shown in Figure 3b. For example, before S401, communication device B sends first information indicating the model alignment capability of communication device B to communication device A. Communication device A determines that the target model alignment mode is a model-based alignment mode based on the model alignment capability of communication device B and communication device A itself. Communication device A then sends a model alignment request to communication device B, and the model alignment request includes second information, i.e., the model alignment request indicates the target model alignment mode. Communication device B then sends a model alignment response to communication device A to confirm the model alignment request of communication device A.

[0128] S403: Communication device B sends the second model to communication device A. Correspondingly, communication device A receives the second model.

[0129] In this embodiment, the target information corresponding to the target model alignment mode is the second model.

[0130] Due to the different capabilities of different communication devices, some communication devices may not support AI models of all structural types. If the model structure type of the model sent by communication device B to communication device A is a model structure type that is not supported by communication device A, communication device A cannot be trained based on the model. Optionally, in order to avoid wasting air interface resources by transmitting invalid models and improve the efficiency of model alignment, before S403, communication device A may also send a third message to communication device B, where the third information indicates one or more model structure types supported by communication device A (hereinafter referred to as candidate model structure types). The candidate model structure types include, for example, at least one of CNN, RNN, and transformer. Communication device B determines the second model based on the candidate model structure type indicated by communication device A in the third information, that is, the model structure type of the second model is a structure type in the candidate model structure type indicated by communication device A.

[0131] In another implementation, the model structure types supported by the communication device A may also be included in the first information, thereby reducing signaling overhead.

[0132] S404: Communication device A trains the first model based on the second model.

[0133] Communication device A trains the first model based on the second model to align the first model with the second model. Communication device A may train the first model based on local training sample data and the second model.

[0134] When the first model is a transmitting model and the second model is a receiving model, communication device A inputs local training sample data into the first model to obtain first output data of the first model. The first output data is then input into the second model to obtain second output data of the second model. Communication device A calculates a first loss value based on the input data and the second output data. This first loss value represents the end-to-end loss. Communication device A calculates a gradient based on this first loss value and then optimizes the first model based on this gradient.

[0135] When the first model is a transmitting model and the second model is also a transmitting model, communication device A inputs local training sample data into the first model as input data to obtain the first output data of the first model, and inputs local training sample data into the second model as input data to obtain the third output data of the second model. Communication device A calculates a second loss value based on the first and third output data. The communication device calculates a gradient based on the second loss value and then optimizes the first model based on the gradient. This results in the performance of the first and second models being close or consistent. Alternatively, communication device A inputs local training sample data into the second model to obtain the third output data of the second model, and then uses the third output data as training samples to train the first model, thereby achieving close or consistent performance of the first and second models.

[0136] Optionally, after the first model is trained for a preset number of rounds or after the first model converges, the communication device A may request the communication device B to verify whether the first model is aligned with the second model.

[0137] S405: Communication device A sends verification information of the first model to communication device B. Correspondingly, communication device B receives the verification information.

[0138] In one possible implementation, the verification information may include model performance information. Model performance information may include, for example, accuracy, precision, or recall. When the accuracy / precision / recall is greater than a model performance threshold, the first model may be considered aligned with the second model; otherwise, the first model and the second model are considered not yet aligned. Alternatively, the verification information may include input data and output data of the first model, and the communication device B verifies whether the end-to-end loss between the first model and the second model is less than a loss threshold. If the loss is less than the loss threshold, the first model and the second model may be considered aligned; otherwise, the first model and the second model are considered not yet aligned.

[0139] S406: Communication device B sends the verification result to communication device A. Correspondingly, communication device A receives the verification result.

[0140] After communication device B determines a verification result based on the verification information, it sends the verification result to communication device A. The verification result indicates that the first model is aligned with the second model, or that the first model is not aligned with the second model. Therefore, if the verification result indicates that the first model is not aligned with the second model, communication device A continues to train the first model to align the first model with the second model.

[0141] When the first model is aligned with the second model, S407 may be optionally further performed.

[0142] S407: Communication device A and communication device B associate the first model with the second model.

[0143] After the first model and the second model are aligned, communication device A and communication device B may interact to associate the first model with the second model.

[0144] For example, the communication device A can configure a first identifier for the first model and a second identifier for the second model. The communication device A associates the first identifier and the second identifier to associate the first model and the second model, and sends the second identifier to the communication device B. The communication device B uses the second identifier as the identifier of the second model. The first identifier and the second identifier can be the same or different. The first identifier of the first model can be configured by the communication device A before the first model is aligned with the second model, or after the first model is aligned with the second model. The communication device A can determine the first identifier of the first model based on the reception time of the model alignment response. Of course, the communication device A can also configure the identifier for the first model according to other principles. For example, a sequence can be randomly generated as the identifier of the first model, or the sequence number of the aligned models can be used as the identifier of the first model. There is no limitation here.

[0145] For another example, communication device A may associate the first model with the second model using the session identifier of the model alignment request or the time of receipt of the model alignment response as a third identifier. Communication device A associates the third identifier with the first model and sends the third identifier to communication device B. Communication device B associates the third identifier with the second model.

[0146] For example, the communication device B may configure a first identifier for the first model and a second identifier for the second model. The communication device B associates the first identifier and the second identifier to associate the first model and the second model, and sends the first identifier to the communication device A. The communication device A uses the first identifier as the identifier of the first model. The first identifier and the second identifier may be the same or different. The second identifier of the second model may be configured by the communication device B before the first model is aligned with the second model, or after the first model is aligned with the second model. For example, the communication device B uses the time when the model alignment request is received as the identifier of the second model. Of course, the communication device B may also assign an identifier to the second model according to other principles. For example, a sequence may be randomly generated as the identifier of the second model, or the sequence of the aligned models may be numbered as the identifier of the second model. This is not limited here.

[0147] The first identifier is associated with the second identifier, or the third identifier is associated with the first model and the second model. This association relationship can be used for model selection / switching / activation / monitoring in subsequent processes. For example, when the first model is subsequently used to process the data to be sent, communication device A determines the second identifier or third identifier of the second model associated with the first model based on the association relationship. Communication device A can indicate the second identifier or third identifier to communication device B to instruct communication device B to process the received signal based on the second model corresponding to the second identifier or third identifier, thereby ensuring end-to-end performance. Alternatively, when communication device B uses the second model to process the data to be sent, communication device B can indicate the second identifier or third identifier to communication device A to instruct communication device A to determine the first model based on the second identifier or third identifier and the association relationship, and use the first model to process the signal from communication device B.

[0148] In this embodiment, when both communicating parties support the model-based alignment mode, the model-based alignment mode can be used for alignment. Using the model-based alignment mode, communication device A can perform offline training on the first model and calculate gradients based on the second model. This eliminates the need for frequent information exchange with communication device B during training, improving training efficiency and reducing air interface resources used during model alignment.

[0149] For the above-mentioned scenario 2, as shown in Figure 5, Figure 5 is a flow chart of another communication method provided by the present application. This embodiment is implemented by communication device A and communication device B. The first model is the model in communication device A, and the second model is the model in communication device B. It should be noted that communication device A can be the first communication device in Figure 2, and communication device B is the second communication device in Figure 2. Alternatively, communication device A can be the second communication device in Figure 2, and communication device B is the first communication device in Figure 2. This embodiment includes the following steps:

[0150] S501: Communication device A sends a model alignment request to communication device B. Correspondingly, communication device B receives the model alignment request.

[0151] S502: Communication device B sends a model alignment response to communication device A. Correspondingly, communication device A receives the model alignment response.

[0152] S503: Communication device B sends a data set to communication device A. Correspondingly, communication device A receives the data set.

[0153] In this embodiment, the target information corresponding to the target model alignment mode is a data set.

[0154] When the first model and the second model are both sending models, the data set may include input data and output data of the second model.

[0155] When the first model is a sending model and the second model is a receiving model, communication device A may send initial sample data to communication device B before S503. Communication device B obtains a data set based on the initial sample data and the third model. The third model is the sending model corresponding to the second model in communication device B. The data set includes output data from the third model.

[0156] S504: Communication device A trains a first model based on the data set.

[0157] Communication device A trains the first model based on the data in the dataset to align the first model with the second model.

[0158] Communication device A inputs the input data in the data set into the first model to obtain fourth output data, calculates a loss value between the fourth output data and the output data in the data set, and optimizes the first model according to the loss value.

[0159] S505: Communication device A sends verification information of the first model to communication device B.

[0160] In this embodiment, the verification information includes, for example, input data and output data of the first model. The input data of the first model is, for example, sample data in the verification set.

[0161] Communication device B uses the output data in the verification information as input data for the second model, obtaining fifth output data from the second model. Communication device B calculates a loss value between the fifth output data and the input data in the verification information, and determines whether the first model and the second model are aligned based on this loss value. This loss value represents an end-to-end loss value. If this loss value is less than an end-to-end loss threshold, the second model is determined to be aligned with the first model; otherwise, they are not aligned.

[0162] S506: Communication device B sends the verification result to communication device A. Correspondingly, communication device A receives the verification result.

[0163] S507: Communication device A and communication device B associate the first model with the second model.

[0164] In this embodiment, when both communicating parties support the dataset-based alignment mode, model alignment can be performed based on the dataset alignment mode. Using the model-based alignment mode, communication device A can perform offline training on the first model, reducing the capability requirements of communication device A. Furthermore, the model alignment process can be completed without disclosing the models of the communication devices.

[0165] For the above-mentioned scenario three, as shown in Figure 6, Figure 6 is a flow chart of another communication method provided by the present application. This embodiment is implemented by communication device A and communication device B. The first model is the model in communication device A, and the second model is the model in communication device B. It should be noted that communication device A can be the first communication device in Figure 2, and communication device B is the second communication device in Figure 2. Alternatively, communication device A can be the second communication device in Figure 2, and communication device B is the first communication device in Figure 2. This embodiment includes the following steps:

[0166] S601: Communication device A sends a model alignment request to communication device B. Correspondingly, communication device B receives the model alignment request.

[0167] S602: Communication device B sends a model alignment response to communication device A. Correspondingly, communication device A receives the model alignment response.

[0168] S603: Communication device A sends output data of the first model to communication device B. Correspondingly, communication device B receives the output data.

[0169] Communication device A inputs training sample data into the first model to obtain output data of the first model. During a training round, communication device A can send the output data of the first model in that training round to communication device B. Communication device B then calculates the reverse gradient (sometimes referred to as the gradient in this application) based on the output data of the first model in that training round.

[0170] S604: Communication device B sends the gradient to communication device A. Correspondingly, communication device A receives the gradient.

[0171] In this embodiment, the target information corresponding to the target model alignment mode is a gradient.

[0172] S605: Communication device A trains the first model based on the gradient.

[0173] Communication device A trains the first model based on the gradient to align the first model with the second model. The gradient refers to the rate of change of the loss function at a point, and its direction points to the direction of the maximum change in the function value. The parameters in the first model are updated based on the gradient, for example, using a gradient descent algorithm to minimize the loss function.

[0174] S606: Communication device A sends verification information of the first model to communication device B.

[0175] In this embodiment, the verification information includes, for example, input data and output data of the first model. The input data of the first model is, for example, sample data in a validation set. When communication device B includes the sample data in the validation set, the verification information may not include the input data of the first model.

[0176] Communication device B uses the output data in the verification information as input data for the second model, obtaining fifth output data from the second model. Communication device B calculates a loss value between the fifth output data and the input data in the verification information, and determines whether the first model and the second model are aligned based on this loss value. This loss value represents an end-to-end loss value. If this loss value is less than an end-to-end loss threshold, the second model is determined to be aligned with the first model; otherwise, they are not aligned.

[0177] S607: Communication device B sends the verification result to communication device A. Correspondingly, communication device A receives the verification result.

[0178] S608: Communication device A and communication device B associate the first model with the second model.

[0179] In this embodiment, when both communicating parties support the gradient-based alignment mode, model alignment can be performed based on the gradient-based alignment mode. Furthermore, during the model alignment process, model alignment can be completed without disclosing the model of the communication device.

[0180] In the embodiment of this application,

[0181] For the above-mentioned scenarios 4 and 5, as shown in FIG7, FIG7 is a flow chart of another communication method provided by the present application. This embodiment is implemented by communication device A and communication device B. The first model is the model in communication device A, and the second model is the model in communication device B. It should be noted that communication device A can be the first communication device in FIG2, and communication device B is the second communication device in FIG2. Alternatively, communication device A can be the second communication device in FIG2, and communication device B is the first communication device in FIG2. This embodiment includes the following steps:

[0182] S701: Communication device A sends a model alignment request to communication device B. Correspondingly, communication device B receives the model alignment request.

[0183] S702: Communication device B sends a model alignment response to communication device A. Correspondingly, communication device A receives the model alignment response.

[0184] S703: Communication device B sends target information corresponding to the target model alignment mode to communication device A. Correspondingly, communication device A receives the target information. The target information includes the second model or data set.

[0185] S704: Communication device A trains the first model based on the second model / data set.

[0186] When the target model alignment mode is the model-based alignment mode, the target information includes the second model. Communication device A trains the first model based on the second model to align the first model with the second model. Communication device A can train the first model based on local training sample data and the second model.

[0187] When the first model is a receiving model and the second model is a transmitting model, communication device A inputs local training sample data into the second model to obtain output data from the second model. The output data from the second model is then input into the first model to obtain output data from the first model. Communication device A calculates a first loss value based on the input data from the second model and the output data from the first model. This first loss value represents the end-to-end loss. Communication device A calculates a gradient based on this first loss value and then optimizes the first model based on this gradient.

[0188] When the first model is the receiving model and the second model is also the receiving model, communication device A inputs local training sample data into the second model to obtain the second model's output data, and also inputs local training sample data into the first model to obtain the first model's output data. Communication device A calculates a loss value based on the second model's output data and the first model's output data. The communication device calculates a gradient based on this loss value and then optimizes the first model based on this gradient. This results in the first model and the second model achieving similar or consistent performance.

[0189] When the target model alignment mode is a dataset-based alignment mode, the target information includes the dataset, which may include input data and output data of the second model.

[0190] After the first model has been trained for a preset number of rounds, or after the first model has converged, communication device A may verify whether the first model is aligned with the second model. Optionally, if the target model alignment mode is a dataset-based alignment mode, communication device A may request communication device B to send verification information, which may include the input and output data of the second model based on sample data in the validation set, so that communication device A can verify whether the first model is aligned with the second model based on the verification information. Optionally, after the verification is successful, i.e., after confirming that the first model is aligned with the second model, communication device A may execute S705.

[0191] S705: Communication device A sends a verification result to communication device B. Correspondingly, communication device B receives the verification result.

[0192] The verification result indicates that the first model is aligned with the second model. When the first model is aligned with the second model, S706 may be optionally further performed.

[0193] Of course, when the communication device A verifies that the first model and the second model are not aligned, it may also send a verification result indicating that the first model and the second model are not aligned to the communication device B.

[0194] S706: Communication device A and communication device B associate the first model with the second model.

[0195] This step is similar to S407. For details, please refer to the relevant description of S407, so it will not be repeated here.

[0196] In this embodiment, when both communication device A and communication device B support model-based alignment mode, the first model and the second model can be aligned based on the model alignment model. Furthermore, when the first model is a receiving model, the first model can be trained and verified based on local training sample data and verification sample data, which can improve the efficiency of model alignment and eliminate the need for frequent interaction with communication device B during the model alignment process, thereby reducing the overhead of the alignment process.

[0197] In the above-mentioned method embodiments of the present application, before the communicating parties perform model alignment, one of the communicating parties informs the other party of its own model alignment capabilities, so that a model alignment mode supported by both communicating parties can be selected for model alignment, so that communication devices with different capabilities, different states or different requirements can flexibly decide on the target model alignment mode, which can improve the efficiency and success rate of model alignment. Furthermore, after the first model and the second model are aligned, the first model and the second model can be associated to associate the first model and the second model, so that in the subsequent process, the model selection / switching / activation / monitoring, etc. can be performed based on the association relationship of the models, which facilitates the use and management of the aligned models.

[0198] Referring to Figure 8 , an embodiment of the present application provides a communication device 800. This communication device 800 can implement the functions of the second communication device or the first communication device in the above-described method embodiment, thereby also achieving the beneficial effects of the above-described method embodiment. In this embodiment of the present application, the communication device 800 can be the first communication device (or the second communication device), or it can be an integrated circuit or component, such as a chip, within the second communication device (or the first communication device).

[0199] It should be noted that the transceiver unit 802 may include a sending unit and a receiving unit, which are respectively used to perform sending and receiving.

[0200] In one possible implementation, when the device 800 is used to execute the method executed by the first communication device in the aforementioned embodiment, the device 800 includes a processing unit 801 and a transceiver unit 802; the transceiver unit 802 is used to send first information, where the first information indicates a model alignment mode supported by the first communication device. The transceiver unit 802 is used to receive second information, where the second information indicates a target model alignment mode, where the target model alignment mode is a model alignment mode supported by the second communication device, and the target alignment mode is a model alignment mode among the model alignment modes supported by the first communication device. The processing unit 801 is used to perform model alignment based on the target model alignment mode.

[0201] In a possible implementation, the model alignment mode includes at least one of the following: a model-based alignment mode; a gradient-based alignment mode; and a dataset-based alignment mode.

[0202] In a possible implementation, the transceiver unit 802 is used to send or receive target information corresponding to the target model alignment mode, and the target information is used to train the first model.

[0203] In one possible implementation, the target alignment mode is a model-based alignment mode, and the target information includes the second model;

[0204] The first model is a model of the first communication device, and the second model is a model of the second communication device; or the first model is a model of the second communication device, and the second model is a model of the first communication device.

[0205] In one possible implementation, the target alignment mode is a gradient-based alignment mode, the target information includes a gradient, and the gradient is obtained based on the output of the first model and the second model; the first model is a sending neural network model, and the second model is a receiving neural network model corresponding to the first model;

[0206] The first model is a model of the first communication device, and the second model is a model of the second communication device; or the first model is a model of the second communication device, and the second model is a model of the first communication device.

[0207] In a possible implementation, the target alignment mode is a data set-based alignment mode, the target information includes a data set, and the data set includes input data and / or output data of the second model;

[0208] The first model is a model of the first communication device, and the second model is a model of the second communication device; or the first model is a model of the second communication device, and the second model is a model of the first communication device.

[0209] In a possible implementation, performing model alignment based on the target model alignment mode further includes:

[0210] The first communication device sends or receives verification information, where the verification information is obtained based on the first model and is used to verify whether the first model is aligned with the second model.

[0211] In a possible implementation, the processing unit 801 is configured to associate the first model with the second model when determining that the first model is aligned with the second model.

[0212] In a possible implementation, the first information further indicates the priority of the model alignment mode supported by the first communication device, and the target model alignment mode is the model alignment mode with the highest priority among the model alignment modes supported by the first communication device and the second communication device.

[0213] In a possible implementation manner, the first information includes at least one of the following: a first bitmap indicating a model alignment mode supported by the first communication device; and an index of the model alignment mode supported by the first communication device.

[0214] In a possible implementation manner, the second information includes at least one of the following: a second bitmap indicating the target model alignment mode; and an index of the target model alignment mode.

[0215] In one possible implementation, when the device 800 is used to execute the method executed by the second communication device in the aforementioned embodiment, the device 800 includes a processing unit 801 and a transceiver unit 802; the transceiver unit 802 is used to receive first information, where the first information indicates a model alignment mode supported by the first communication device. The transceiver unit 802 is used to send second information, where the second information indicates a target model alignment mode, where the target model alignment mode is a model alignment mode supported by the second communication device, and the target alignment mode is a model alignment mode among the model alignment modes supported by the first communication device. The processing unit 801 is used to perform model alignment based on the target model alignment mode.

[0216] In a possible implementation, the model alignment mode includes at least one of the following: a model-based alignment mode; a gradient-based alignment mode; and a dataset-based alignment mode.

[0217] In one possible implementation, the second communication device performs model alignment based on the target model alignment mode, including: the second communication device sends or receives target information corresponding to the target model alignment mode, and the target information is used to train the first model.

[0218] In one possible implementation, the target alignment mode is a model-based alignment mode, and the target information includes a second model; wherein the first model is a model of the first communication device, and the second model is a model of the second communication device; or, the first model is a model of the second communication device, and the second model is a model of the first communication device.

[0219] In one possible implementation, the target alignment mode is a gradient-based alignment mode, the target information includes a gradient, and the gradient is obtained based on the output of the first model and the second model; the first model is a sending neural network model, and the second model is a receiving neural network model corresponding to the first model;

[0220] The first model is a model of the first communication device, and the second model is a model of the second communication device; or the first model is a model of the second communication device, and the second model is a model of the first communication device.

[0221] In a possible implementation, the target alignment mode is a data set-based alignment mode, the target information includes a data set, and the data set includes input data and / or output data of the second model;

[0222] The first model is a model of the first communication device, and the second model is a model of the second communication device; or the first model is a model of the second communication device, and the second model is a model of the first communication device.

[0223] In a possible implementation, the transceiver unit 802 is further configured to send or receive verification information, where the verification information is obtained based on the first model and is used to verify whether the first model is aligned with the second model.

[0224] In a possible implementation, the processing unit 801 is configured to associate the first model with the second model when the first model is aligned with the second model.

[0225] In a possible implementation, the first information further indicates the priority of the model alignment mode supported by the first communication device, and the target model alignment mode is the model alignment mode with the highest priority among the model alignment modes supported by the first communication device and the second communication device.

[0226] In a possible implementation manner, the first information includes at least one of the following: a first bitmap indicating a model alignment mode supported by the first communication device; and an index of the model alignment mode supported by the first communication device.

[0227] In a possible implementation manner, the second information includes at least one of the following: a second bitmap indicating the target model alignment mode; and an index of the target model alignment mode.

[0228] It should be noted that, for details on the information execution process of the units of the above-mentioned communication device 800, please refer to the description in the method embodiment shown above in this application, and no further details will be given here.

[0229] Please refer to Figure 9, which is another schematic structural diagram of a communication device 900 provided in this application. The communication device 900 includes a logic circuit 901 and an input / output interface 902. The communication device 900 may be a chip or an integrated circuit.

[0230] The transceiver unit 802 shown in FIG8 may be a communication interface, which may be the input / output interface 902 in FIG9 , which may include an input interface and an output interface. Alternatively, the communication interface may be a transceiver circuit, which may include an input interface circuit and an output interface circuit.

[0231] Optionally, the logic circuit 901 is used to determine first information, where the first information indicates a model alignment mode supported by the first communication device; and the input / output interface 902 is used to send the first information.

[0232] Optionally, the input-output interface 902 is used to receive second information, wherein the second information indicates a target model alignment mode, the target model alignment mode is a model alignment mode supported by the second communication device, and the target alignment mode is a model alignment mode among the model alignment modes supported by the first communication device; the logic circuit 901 is used to perform model alignment based on the target model alignment mode.

[0233] The logic circuit 901 and the input / output interface 902 may also execute other steps executed by the first communication device or the second communication device in any embodiment and achieve corresponding beneficial effects, which will not be described in detail here.

[0234] In a possible implementation, the processing unit 801 shown in FIG. 8 may be the logic circuit 901 in FIG. 9 .

[0235] Optionally, the logic circuit 901 may be a processing device, and the functions of the processing device may be partially or entirely implemented by software. The functions of the processing device may be partially or entirely implemented by software.

[0236] Optionally, the processing device may include a memory and a processor, wherein the memory is used to store a computer program, and the processor reads and executes the computer program stored in the memory to perform corresponding processing and / or steps in any one of the method embodiments.

[0237] Alternatively, the processing device may include only a processor. A memory for storing the computer program is located outside the processing device, and the processor is connected to the memory via circuits / wires to read and execute the computer program stored in the memory. The memory and processor may be integrated or physically separate.

[0238] Optionally, the processing device may be one or more chips, or one or more integrated circuits. For example, the processing device may be one or more field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), system-on-chips (SoCs), central processor units (CPUs), network processors (NPs), digital signal processors (DSPs), microcontroller units (MCUs), programmable logic devices (PLDs), or other integrated chips, or any combination of the above chips or processors.

[0239] Please refer to Figure 10, which shows the communication device 1000 involved in the above-mentioned embodiments provided in an embodiment of the present application. The communication device 1000 can specifically be a communication device serving as a terminal device in the above-mentioned embodiments. The example shown in Figure 10 is that the terminal device is implemented through the terminal device (or a component in the terminal device).

[0240] Herein, a possible logical structure diagram of the communication device 1000 is shown. The communication device 1000 may include but is not limited to at least one processor 1001 and a communication port 1002 .

[0241] The transceiver unit 802 shown in FIG8 may be a communication interface, which may be the communication port 1002 in FIG10 , which may include an input interface and an output interface. Alternatively, the communication port 1002 may be a transceiver circuit, which may include an input interface circuit and an output interface circuit.

[0242] Further optionally, the device may also include at least one of a memory 1003 and a bus 1004. In an embodiment of the present application, the at least one processor 1001 is used to control and process the actions of the communication device 1000.

[0243] In addition, the processor 1001 can be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field programmable gate array or other programmable logic device, a transistor logic device, a hardware component, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. The processor can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a digital signal processor and a microprocessor, and so on. Those skilled in the art will clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0244] It should be noted that the communication device 1000 shown in Figure 10 can be specifically used to implement the steps implemented by the terminal device in the aforementioned method embodiment and achieve the corresponding technical effects of the terminal device. The specific implementation methods of the communication device shown in Figure 10 can refer to the description in the aforementioned method embodiment and will not be repeated here.

[0245] Please refer to Figure 11, which is a structural diagram of the communication device 1100 involved in the above-mentioned embodiments provided in an embodiment of the present application. The communication device 1100 can specifically be a communication device as a network device in the above-mentioned embodiments. The example shown in Figure 11 is that the network device is implemented through the network device (or a component in the network device), wherein the structure of the communication device can refer to the structure shown in Figure 11.

[0246] The communication device 1100 includes at least one processor 1111 and at least one network interface 1114. Further optionally, the communication device also includes at least one memory 1112, at least one transceiver 1113 and one or more antennas 1115. The processor 1111, the memory 1112, the transceiver 1113 and the network interface 1114 are connected, for example, via a bus. In an embodiment of the present application, the connection may include various interfaces, transmission lines or buses, etc., which are not limited in this embodiment. The antenna 1115 is connected to the transceiver 1113. The network interface 1114 is used to enable the communication device to communicate with other communication devices through a communication link. For example, the network interface 1114 may include a network interface between the communication device and the core network device, such as an S1 interface, and the network interface may include a network interface between the communication device and other communication devices (such as other network devices or core network devices), such as an X2 or Xn interface.

[0247] The transceiver unit 802 shown in FIG8 may be a communication interface, which may be the network interface 1114 in FIG11 , which may include an input interface and an output interface. Alternatively, the network interface 1114 may be a transceiver circuit, which may include an input interface circuit and an output interface circuit.

[0248] Processor 1111 is primarily used to process communication protocols and communication data, control the entire communication device, execute software programs, and process software program data, for example, to support the communication device in performing the actions described in the embodiments. The communication device may include a baseband processor and a central processing unit. The baseband processor is primarily used to process communication protocols and communication data, while the central processing unit is primarily used to control the entire terminal device, execute software programs, and process software program data. Processor 1111 in Figure 11 may integrate the functions of both a baseband processor and a central processing unit. Those skilled in the art will appreciate that the baseband processor and the central processing unit may also be independent processors interconnected via a bus or other technology. Those skilled in the art will appreciate that a terminal device may include multiple baseband processors to accommodate different network standards, multiple central processing units to enhance its processing capabilities, and various components of the terminal device may be connected via various buses. The baseband processor may also be referred to as a baseband processing circuit or a baseband processing chip. The central processing unit may also be referred to as a central processing circuit or a central processing chip. The functionality for processing communication protocols and communication data may be built into the processor or stored in memory as a software program, which is executed by the processor to implement the baseband processing functionality.

[0249] The memory is primarily used to store software programs and data. Memory 1112 can exist independently and be connected to processor 1111. Alternatively, memory 1112 can be integrated with processor 1111, for example, within a single chip. Memory 1112 can store program code for executing the technical solutions of the embodiments of the present application, and execution is controlled by processor 1111. The various computer program codes executed can also be considered drivers for processor 1111.

[0250] Figure 11 shows only one memory and one processor. In an actual terminal device, there may be multiple processors and multiple memories. The memory may also be referred to as a storage medium or a storage device. The memory may be a storage element on the same chip as the processor, i.e., an on-chip storage element, or an independent storage element, which is not limited in the present embodiment.

[0251] The transceiver 1113 can be used to support the reception or transmission of radio frequency signals between the communication device and the terminal. The transceiver 1113 can be connected to the antenna 1115. The transceiver 1113 includes a transmitter Tx and a receiver Rx. Specifically, one or more antennas 1115 can receive radio frequency signals. The receiver Rx of the transceiver 1113 is used to receive the radio frequency signal from the antenna, convert the radio frequency signal into a digital baseband signal or a digital intermediate frequency signal, and provide the digital baseband signal or digital intermediate frequency signal to the processor 1111 so that the processor 1111 can further process the digital baseband signal or digital intermediate frequency signal, such as demodulation and decoding. In addition, the transmitter Tx in the transceiver 1113 is also used to receive a modulated digital baseband signal or digital intermediate frequency signal from the processor 1111, convert the modulated digital baseband signal or digital intermediate frequency signal into a radio frequency signal, and transmit the radio frequency signal through one or more antennas 1115. Specifically, the receiver Rx can selectively perform one or more stages of down-mixing and analog-to-digital conversion on the RF signal to obtain a digital baseband signal or a digital intermediate frequency signal. The order of the down-mixing and analog-to-digital conversion processes is adjustable. The transmitter Tx can selectively perform one or more stages of up-mixing and digital-to-analog conversion on the modulated digital baseband signal or digital intermediate frequency signal to obtain a RF signal. The order of the up-mixing and digital-to-analog conversion processes is adjustable. The digital baseband signal and the digital intermediate frequency signal may be collectively referred to as digital signals.

[0252] The transceiver 1113 may also be referred to as a transceiver unit, a transceiver, a transceiver device, etc. Optionally, a device in the transceiver unit that implements a receiving function may be referred to as a receiving unit, and a device in the transceiver unit that implements a transmitting function may be referred to as a transmitting unit. That is, the transceiver unit includes a receiving unit and a transmitting unit. The receiving unit may also be referred to as a receiver, an input port, a receiving circuit, etc., and the transmitting unit may be referred to as a transmitter, a transmitter, or a transmitting circuit, etc.

[0253] It should be noted that the communication device 1100 shown in Figure 11 can be specifically used to implement the steps implemented by the network device in the aforementioned method embodiment, and to achieve the corresponding technical effects of the network device. The specific implementation methods of the communication device 1100 shown in Figure 11 can refer to the description in the aforementioned method embodiment, and will not be repeated here.

[0254] Please refer to FIG12 , which is a schematic structural diagram of the communication device involved in the above-mentioned embodiment provided in an embodiment of the present application.

[0255] It can be understood that the communication device 120 includes, for example, modules, units, elements, circuits, or interfaces, which are appropriately configured together to implement the technical solutions provided in this application. The communication device 120 can be the terminal device or network device described above, or a component (such as a chip) in these devices, used to implement the method described in the following method embodiment. The communication device 120 includes one or more processors 121. The processor 121 can be a general-purpose processor or a dedicated processor. For example, it can be a baseband processor or a central processing unit. The baseband processor can be used to process communication protocols and communication data, and the central processing unit can be used to control the communication device (such as a RAN node, terminal, or chip, etc.), execute software programs, and process data of software programs.

[0256] Optionally, in one design, the processor 121 may include a program 123 (sometimes also referred to as code or instructions), which may be executed on the processor 121 to cause the communication device 120 to perform the methods described in the following embodiments. In yet another possible design, the communication device 120 includes circuitry (not shown in FIG12 ).

[0257] Optionally, the communication device 120 may include one or more memories 122 on which a program 124 (sometimes also referred to as code or instructions) is stored. The program 124 can be run on the processor 121, so that the communication device 120 executes the method described in the above method embodiment.

[0258] Optionally, the processor 121 and / or the memory 122 may include an AI module 127, 128, which is used to implement AI-related functions. The AI ​​module may be implemented through software, hardware, or a combination of software and hardware. For example, the AI ​​module may include a wireless intelligent control (RIC) module. For example, the AI ​​module may be a near real-time RIC or a non-real-time RIC.

[0259] Optionally, data may be stored in the processor 121 and / or the memory 122. The processor and the memory may be provided separately or integrated together.

[0260] Optionally, the communication device 120 may further include a transceiver 125 and / or an antenna 126. The processor 121 may also be sometimes referred to as a processing unit, and controls the communication device (e.g., a RAN node or terminal). The transceiver 125 may also be sometimes referred to as a transceiver unit, a transceiver, a transceiver circuit, or a transceiver, and is configured to implement the transceiver functions of the communication device through the antenna 126.

[0261] The transceiver unit 802 shown in FIG8 may be a communication interface, which may be the transceiver 125 in FIG12 . The transceiver 125 may include an input interface and an output interface. Alternatively, the transceiver 125 may be a transceiver circuit, which may include an input interface circuit and an output interface circuit.

[0262] An embodiment of the present application further provides a computer-readable storage medium, which is used to store one or more computer-executable instructions. When the computer-executable instructions are executed by a processor, the processor executes the method described in the possible implementation methods of the first communication device or the second communication device in the aforementioned embodiment.

[0263] An embodiment of the present application also provides a computer program product (or computer program). When the computer program product is executed by the processor, the processor executes the method that may be implemented by the above-mentioned first communication device or second communication device.

[0264] An embodiment of the present application also provides a chip system, which includes at least one processor for supporting a communication device to implement the functions involved in the possible implementation methods of the above-mentioned communication device. Optionally, the chip system also includes an interface circuit, which provides program instructions and / or data to the at least one processor. In one possible design, the chip system may also include a memory, which is used to store the necessary program instructions and data for the communication device. The chip system can be composed of chips, or it can include chips and other discrete devices, wherein the communication device can specifically be the first communication device or the second communication device in the aforementioned method embodiment.

[0265] An embodiment of the present application further provides a communication system, wherein the network system architecture includes the first communication device and the second communication device in any of the above embodiments.

[0266] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0267] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0268] In addition, the functional units in the various embodiments of the present application can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit. If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the contributing part or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

Claims

1. A communication method, characterized in that: The method comprises: A first communication device sends first information, where the first information indicates a model alignment mode supported by the first communication device; The first communication device receives second information, the second information indicates a target model alignment mode, the target model alignment mode is a model alignment mode supported by the second communication device, and the target alignment mode is a model alignment mode among the model alignment modes supported by the first communication device; The first communication device performs model alignment based on the target model alignment mode.

2. The method according to claim 1, characterized in that The model alignment mode includes at least one of the following: Model-based alignment mode; Gradient-based alignment mode; Alignment mode based on the dataset.

3. The method according to claim 1 or 2, characterized in that: The first communication device performs model alignment based on the target model alignment mode, including: The first communication device sends or receives target information corresponding to the target model alignment mode, and the target information is used to train the first model.

4. The method according to claim 3, characterized in that: The target alignment mode is a model-based alignment mode, and the target information includes a second model; The first model is a model of the first communication device, and the second model is a model of the second communication device; or the first model is a model of the second communication device, and the second model is a model of the first communication device.

5. The method according to claim 3, characterized in that: The target alignment mode is a gradient-based alignment mode, the target information includes a gradient, and the gradient is obtained according to the output of the first model and the second model; the first model is a sending neural network model, and the second model is a receiving neural network model corresponding to the first model; The first model is a model of the first communication device, and the second model is a model of the second communication device; or the first model is a model of the second communication device, and the second model is a model of the first communication device.

6. The method according to claim 3, characterized in that The target alignment mode is a data set-based alignment mode, the target information includes a data set, and the data set includes input data and / or output data of the second model; The first model is a model of the first communication device, and the second model is a model of the second communication device; or the first model is a model of the second communication device, and the second model is a model of the first communication device.

7. The method according to any one of claims 3 to 6, characterized in that The performing model alignment based on the target model alignment mode further includes: The first communication device sends or receives verification information, where the verification information is obtained according to the first model, and the verification information is used to verify whether the first model is aligned with the second model.

8. The method according to claim 7, characterized in that The method further comprises: If the first model is aligned with the second model, the first communication device associates the first model with the second model.

9. The method according to any one of claims 1 to 8, characterized in that The first information further indicates the priority of the model alignment mode supported by the first communication device, and the target model alignment mode is the model alignment mode with the highest priority among the model alignment modes supported by the first communication device and the second communication device.

10. The method according to any one of claims 1 to 9, characterized in that The first information includes at least one of the following: a first bitmap indicating a model alignment mode supported by the first communication device; an index of a model alignment mode supported by the first communication device; and / or The second information includes at least one of the following: a second bitmap indicating an alignment mode of the target model; The index of the target model alignment mode.

11. A communication method, characterized in that: The method comprises: The second communication device receives first information, wherein the first information indicates a model alignment mode supported by the first communication device; The second communication device sends second information, where the second information indicates a target model alignment mode, where the target model alignment mode is a model alignment mode supported by the second communication device, and the target alignment mode is a model alignment mode among the model alignment modes supported by the first communication device; The second communication device performs model alignment based on the target model alignment mode.

12. The method according to claim 11, characterized in that The model alignment mode includes at least one of the following: Model-based alignment mode; Gradient-based alignment mode; Alignment mode based on the dataset.

13. The method according to claim 11 or 12, characterized in that: The second communication device performs model alignment based on the target model alignment mode, including: The second communication device sends or receives target information corresponding to the target model alignment mode, and the target information is used to train the first model.

14. The method according to claim 13, characterized in that The target alignment mode is a model-based alignment mode, and the target information includes a second model; The first model is a model of the first communication device, and the second model is a model of the second communication device; or the first model is a model of the second communication device, and the second model is a model of the first communication device.

15. The method according to claim 13, characterized in that The target alignment mode is a gradient-based alignment mode, the target information includes a gradient, and the gradient is obtained according to the output of the first model and the second model; the first model is a sending neural network model, and the second model is a receiving neural network model corresponding to the first model; The first model is a model of the first communication device, and the second model is a model of the second communication device; or the first model is a model of the second communication device, and the second model is a model of the first communication device.

16. The method according to claim 13, characterized in that The target alignment mode is a data set-based alignment mode, the target information includes a data set, and the data set includes input data and / or output data of the second model; The first model is a model of the first communication device, and the second model is a model of the second communication device; or the first model is a model of the second communication device, and the second model is a model of the first communication device.

17. The method according to any one of claims 13 to 16, characterized in that The second communication device performs model alignment based on the target model alignment mode, further comprising: The second communication device sends or receives verification information, where the verification information is obtained according to the first model, and the verification information is used to verify whether the first model is aligned with the second model.

18. The method according to claim 17, characterized in that The method further comprises: If the first model is aligned with the second model, the second communication device associates the first model with the second model.

19. The method according to any one of claims 11 to 18, characterized in that The first information further indicates the priority of the model alignment mode supported by the first communication device, and the target model alignment mode is the model alignment mode with the highest priority among the model alignment modes supported by the first communication device and the second communication device.

20. The method according to any one of claims 11 to 19, characterized in that The first information includes at least one of the following: a first bitmap indicating a model alignment mode supported by the first communication device; an index of a model alignment mode supported by the first communication device; and / or The second information includes at least one of the following: a second bitmap indicating an alignment mode of the target model; The index of the target model alignment mode.

21. A communication device, characterized in that: Comprising means for performing the method as claimed in any one of claims 1 to 20.

22. A communication device, characterized in that: The method comprises at least one processor coupled to a memory; the at least one processor is configured to execute the method according to any one of claims 1 to 20.

23. The communication device according to claim 22, characterized in that The communication device is a chip or a chip system.

24. A readable storage medium, characterized in that: The storage medium stores a computer program or an instruction, and when the computer program or the instruction is executed by the communication device, the method according to any one of claims 1 to 20 is implemented.

25. A computer program product, characterized in that The method comprises instructions which, when executed on a computer, cause the computer to perform the method according to any one of claims 1 to 20.

Images