Communication method and apparatus

By detecting the performance of some AI functions to infer the performance of multiple AI functions, and by utilizing the performance correlation, the problem of high signaling overhead in AI model performance testing is solved, achieving resource saving and management simplification.

WO2026233096A1PCT designated stage Publication Date: 2026-11-12HUAWEI TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/CN2026/091804
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2025-05-08
Filing Date
2026-04-20
Publication Date
2026-11-12

AI Technical Summary

Technical Problem

In existing technologies, the signaling overhead for AI model performance testing is relatively large, leading to resource waste and increased management complexity.

Method used

By testing the performance of some AI functions, the performance of multiple AI functions can be inferred, and the correlation between performance can be used to reduce testing resources and signaling overhead.

Benefits of technology

It reduces the resource and signaling overhead of AI function performance testing, simplifies management complexity, and improves testing efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN2026091804_12112026_PF_FP_ABST
    Figure CN2026091804_12112026_PF_FP_ABST
Patent Text Reader

Abstract

A communication method and apparatus, which are applicable to performance detection of AI functions. The method comprises: a terminal device sending first information; and a network device acquiring the performance of a first AI function, and determining the performance of a second AI function on the basis of the first information and the performance of the first AI function, wherein the first information is used for indicating that there is an association relationship between the performance of the first AI function and the performance of the second AI function. In the method, a network device can infer the performance of another AI function (e.g., a second AI function) on the basis of the performance of one AI function (e.g., a first AI function), and thus can know the performances of a plurality of AI functions simply by means of detecting the performances of some of the plurality of AI functions without detecting the performances of the plurality of AI functions one by one, thereby saving on resource overheads and signaling overheads for detecting the performances of the AI functions.
Need to check novelty before this filing date? Find Prior Art

Description

A communication method and apparatus

[0001] Cross-references to related applications

[0002] This application claims priority to Chinese Patent Application No. 202510592824.0, filed on May 8, 2025, entitled "A Communication Method and Apparatus", the entire contents of which are incorporated herein by reference. Technical Field

[0003] This application relates to the field of communication technology, and in particular to a communication method and apparatus. Background Technology

[0004] Currently, the application of artificial intelligence (AI) in communications, especially in air communication, has been extensively researched. For example, research projects have been initiated focusing on air communication AI, exploring its basic framework and several typical use cases (AI channel state information (CSI) feedback, AI beam management, and AI positioning). Considering that future communication systems will be further integrated with AI technology, many traditional communication modules can be replaced by AI methods. AI methods essentially replace functions implemented using traditional methods with AI functionalities. AI functions are implemented through AI models, which can be closely integrated with specific application scenarios and conditions. For example, AI models can be trained specifically based on application scenarios or conditions, thus achieving superior performance.

[0005] When using an AI model to implement a function, the performance of that function is closely related to the performance of the AI ​​model. To ensure optimal system performance, the performance of the AI ​​model is usually tested. Traditional methods for testing the performance of AI models suffer from significant signaling overhead. Summary of the Invention

[0006] This application provides a communication method and apparatus for reducing the signaling overhead of performance testing of AI functions.

[0007] In a first aspect, a communication method is provided, which is applied to a first device, which may be a terminal-side device (hereinafter referred to as a terminal device) or a network-side device (hereinafter referred to as a network device). The terminal device may be applied to (or be) a terminal device or a component of a terminal device; or applied to (or be) a communication module and / or computing module of a terminal device; or applied to (or be) a circuit or chip in a terminal device responsible for communication functions (such as a modem chip, also known as a baseband chip, or a system-on-chip (SoC) chip containing a modem core or a system-in-package (SIP) chip); or applied to (or be) a circuit or chip in a terminal device responsible for communication and / or computing functions (such as a central processing unit (CPU), digital signal processing (DSP), graphics processing unit (GPU), neural processing unit (NPU), artificial intelligence (AI) processor, application-specific integrated circuit (ASIC), or field-programmable gate array (FPGA)); or applied to a logic node, logic module, or software that can implement all or part of the functions of the terminal device. The network device can be applied to (or be) a network device or a component of a network device; or to (or be) a communication module and / or computing module of a network device; or to (or be) a circuit or chip in a network device responsible for communication functions (such as a modem chip (also known as a baseband chip), or a SoC chip or SIP chip containing a modem core), a chip system, or a processor); or to (or be) a circuit or chip in a network device responsible for communication and / or computing functions (such as a CPU, GPU, NPU, AI processor, ASIC, or FPGA); or to a logical node, logical module, or software that can implement all or part of the functions of a network device. In one example, the method can be applied to one or more of the central unit (CU), distributed unit (DU), and radio unit (RU) of a network device.

[0008] For ease of description, the following example uses the method provided in the first aspect applied to a network device. The method includes: receiving first information to obtain the performance of a first AI function; and determining the performance of a second AI function based on the first information and the performance of the first AI function. The first information indicates a correlation between the performance of the first AI function and the performance of the second AI function.

[0009] In this method, network devices can infer the performance of one AI function (e.g., a second AI function) based on the performance of AI functions with performance correlations. When the performance of multiple AI functions needs to be tested, and these multiple AI functions include those with performance correlations, only the performance of a subset of these AI functions needs to be tested, eliminating the need to test the performance of each individual AI function. This method allows network devices to obtain the performance of more AI functions by testing the performance of fewer AI functions, thereby saving resource and signaling overhead associated with AI function performance testing.

[0010] In a design, the performance of two AI functions is correlated if they satisfy one or more of the following: the performance of one AI function is superior to or equal to the performance of the other AI function; or, the applicable conditions of the two AI functions are correlated. The applicable conditions include one or more of the following: applicable environmental conditions, the capabilities of the device deploying the AI ​​function, the state of the device deploying the AI ​​function, or the associated configuration of the AI ​​function.

[0011] In this design, the applicability of AI functions can be correlated, which can be applied to determining the applicability of inactive AI functions. Alternatively, the performance of two AI functions can be correlated, which can also be applied to detecting the performance of activated AI functions. This design only requires maintaining one mechanism to determine the applicability and performance of AI functions, eliminating the need to maintain separate mechanisms for determining applicability and performance, thus reducing the management complexity of AI functions.

[0012] In one design, two AI functions are designated as a first AI function and a second AI function. The applicable conditions of the two AI functions are related, including: the applicable conditions of a first performance of the first AI function and the applicable conditions of a second performance of the second AI function are related; or, the applicable conditions of a first performance index of the first performance of the first AI function and the applicable conditions of a second performance index of the second performance of the second AI function are related.

[0013] In this design, the applicability conditions of AI functions include the applicability conditions of performance and / or the applicability conditions of performance indicators, which enables more refined applicability testing and reduces the failure of AI functions due to inapplicability.

[0014] There are several designs for the first information. The following are some possible designs for the first information. These designs can coexist, which is quite flexible and helps to manage AI functions flexibly.

[0015] In Design 1, the first information includes at least one set of IDs, wherein one set includes IDs of AI functions associated with performance.

[0016] In this design, AI functions associated with performance are represented by a set of IDs, which facilitates interpretation by network devices.

[0017] Optionally, the first information may also include performance information associated with one or more ID sets. The performance information associated with an ID set may be performance information associated with some or all of the IDs in the ID set. For example, if at least one ID set includes a first ID set, the first information may also include performance information associated with some or all of the IDs in the first ID set.

[0018] In this design, the ID of the AI ​​function is also associated with specific performance information, which can indicate the specific performance that is related between AI functions, thereby enabling granular management of AI functions based on performance type.

[0019] In Design 2, the first information includes at least one AI function ID and at least one set of IDs, wherein one set of IDs corresponds to one ID, including the IDs of AI functions associated with the performance of the AI ​​function indicated by the corresponding ID.

[0020] In this design, an AI function ID corresponds to a set of IDs, which is suitable for scenarios where multiple AI functions may be related to the performance of the same AI function, but the performance of these multiple AI functions is not related.

[0021] Optionally, the first information may also include performance information associated with one or more sets of IDs. The performance information associated with a set of IDs may be performance information associated with some or all of the IDs in the set. For example, at least one AI function ID includes a first ID, at least one ID set includes a first ID set, the first ID and the first ID set correspond, and the first information may also include performance information associated with the first ID, and / or performance information associated with some or all of the IDs in the first ID set.

[0022] In this design, the ID of each AI function can be associated with performance information. This design enables granular management of AI functions based on performance type.

[0023] In Design 3, the first information includes: the ID of at least one AI function, and at least one performance information associated with some or all of the IDs of the at least one AI function.

[0024] In this design, each AI function's ID can be associated with performance information, indicating the correlation between different performance characteristics of different AI functions. This design enables granular management of AI functions based on performance type.

[0025] In Design 1 to Design 3, the first information may also include: at least one performance metric associated with each performance.

[0026] In this design, each performance metric can also be associated with a performance indicator, which can indicate that there is a correlation between AI functions when the performance reaches a certain performance indicator, thereby enabling granular management of AI functions based on performance indicators.

[0027] In one design, the method further includes: determining a management strategy for the second AI function based on its performance, the management strategy including one or more of the following: activating the second AI function, deactivating the second AI function, updating the second AI function, retraining the second AI function, or switching the second AI function.

[0028] In this design, network devices can determine the management strategy for the second AI function based on its performance, which helps improve network performance. For example, for second AI functions that are no longer applicable or have poor performance, there is no need to activate them, thus reducing unnecessary activation of AI functions.

[0029] In one design, the second AI function is activated or its configuration is enabled when the performance of the second AI function meets preset performance indicators.

[0030] In this design, the network device will only activate the second AI function if the performance of the second function meets the actual performance requirements of the terminal device, thereby reducing the activation of unnecessary AI functions.

[0031] In one design, the method further includes: receiving second information, and deactivating a second AI function based on the second information. The second information is used to indicate that a first performance characteristic of the second AI function does not meet a first performance metric.

[0032] In this design, network devices can determine, based on feedback from terminal devices (such as second information), that the first performance of the second AI function does not meet the first performance indicator, and thus deactivate the second AI function, thereby reducing unnecessary AI function operation and conserving network resources as much as possible. This design helps network devices make decisions on management strategies for the second AI function.

[0033] In one design, the method further includes: sending a first request for an AI function whose performance is correlated.

[0034] In this design, when a network device needs to know about AI functions with performance correlations, it sends a first request to trigger the terminal device to report the first information. This design reduces unnecessary reporting of performance-correlated AI functions, thus saving signaling overhead. Furthermore, it allows network devices to obtain the latest performance-correlated AI functions more promptly, enabling a more accurate determination of the performance of one or more AI functions. For example, if a terminal device periodically reports performance-correlated AI functions, it receives the first request before the next period begins and reports the relevant AI functions.

[0035] Secondly, a communication method is provided, which is applied to a second device, which may be a terminal-side device (hereinafter referred to as a terminal device) or a network-side device (hereinafter referred to as a network device). Optionally, when the method of the first aspect is applied to the terminal device, the method of the second aspect is applied to the network device; and when the method of the first aspect is applied to the network device, the method of the second aspect is applied to the terminal device. For a detailed description of the terminal device and the network device, please refer to the description of the terminal device and the network device in the aforementioned first aspect, which will not be repeated here.

[0036] For ease of description, the following example uses the method provided in the second aspect applied to a terminal device. The method includes: determining first information and sending the first information, wherein the first information is used to indicate a correlation between the performance of a first AI function and the performance of a second AI function.

[0037] In a design, the performance of two AI functions is correlated if they satisfy one or more of the following: the performance of one AI function is superior to or equal to the performance of the other AI function; or, the applicable conditions of the two AI functions are correlated. The applicable conditions include one or more of the following: applicable environmental conditions, the capabilities of the device deploying the AI ​​function, the state of the device deploying the AI ​​function, or the associated configuration of the AI ​​function.

[0038] In one design, two AI functions are designated as a first AI function and a second AI function. The applicable conditions of the two AI functions are related, including: the applicable conditions of a first performance of the first AI function and the applicable conditions of a second performance of the second AI function are related; or, the applicable conditions of a first performance index of the first performance of the first AI function and the applicable conditions of a second performance index of the second performance of the second AI function are related.

[0039] In one design, the first information includes at least one set of identifier IDs, wherein one set includes IDs of AI functions associated with performance.

[0040] Optionally, at least one ID set includes a first ID set, and the first information also includes performance information associated with some or all of the IDs in the first ID set.

[0041] In one design, the first information includes at least one AI function ID and at least one set of IDs, wherein one set of IDs corresponds to one ID, including the IDs of AI functions associated with the performance of the AI ​​function indicated by the corresponding ID.

[0042] Optionally, at least one AI function ID includes a first ID, at least one ID set includes a first ID set, the first ID and the first ID set correspond, and the first information also includes performance information associated with the first ID and / or performance information associated with some or all of the IDs in the first ID set.

[0043] In one design, the first information includes the ID of at least one AI function and the ID of at least one performance associated with some or all of the IDs of the at least one AI function.

[0044] In one design, the first information also includes at least one performance metric associated with each performance metric.

[0045] In one design, the method further includes: receiving a first request for requesting AI functions with performance correlation.

[0046] The beneficial effects of the second aspect and its various designs can be referred to the aforementioned beneficial effects of the first aspect and its various designs, and will not be repeated here.

[0047] Thirdly, a communication device is provided for performing the method described in the first or second aspect and any of its designs. The beneficial effects can be found in the relevant descriptions of the first or second aspect, which will not be repeated here.

[0048] In one possible design, the communication device includes corresponding means, modules, or units for performing the methods of the first or second aspect. These modules, units, or means can be implemented in software, hardware, or a combination of both. For example, the communication device includes a processing unit (sometimes also called a processing module or processor) and / or input / output interfaces. Input / output interfaces include input interfaces and / or output interfaces, which can be interface circuits, output circuits, input circuits, pins, or related circuits, etc. Optionally, the communication device also includes a transceiver unit (sometimes also called a transceiver module or transceiver). The transceiver unit is capable of implementing both transmitting and receiving functions. When the transceiver module implements the transmitting function, it can be called a transmitting unit (sometimes also called a transmitting module); when the transceiver module implements the receiving function, it can be called a receiving unit (sometimes also called a receiving module). The transmitting unit and the receiving unit can be the same functional module; or, the transmitting unit and the receiving unit can be different functional modules. In one example, the processing unit includes (or is) a baseband device, and the transceiver unit includes (or is) a radio frequency device. These input / output interfaces and units (modules) can perform the corresponding functions in the method examples of the first or second aspect above. For details, please refer to the detailed description in the method examples, which will not be repeated here.

[0049] For example, when the communication device is used to implement the corresponding function in the method example of the first aspect, the transceiver unit is used to receive first information, which is used to indicate that the performance of the first AI function is related to the performance of the second AI function; the processing unit is used to obtain the performance of the first AI function and determine the performance of the second AI function based on the first information and the performance of the first AI function.

[0050] For example, when the communication device is used to implement the corresponding function in the method example of the second aspect, the processing unit is used to determine first information, which indicates that the performance of the first AI function is related to the performance of the second AI function; the transceiver unit is used to send the first information.

[0051] Fourthly, a communication device is provided, comprising a communication interface and one or more processors. Optionally, the communication interface is an interface circuit. The communication interface is used to implement communication functions within the communication device and / or for the communication device to communicate with other devices or components. In possible designs, one or more processors communicate with other devices or components through the communication interface.

[0052] One or more processors are coupled to a memory for storing part or all of a computer program (also referred to as code or instructions) necessary for implementing the functions involved in the first or second aspect and any of their designs. The one or more processors can execute the computer program (or code or instructions), which, when executed, causes the communication device to implement the methods of the first or second aspect and any of their designs. This application does not limit the specific type of processor. For example, the processor can be a baseband device, a CPU, or other specific integrated circuits. As another example, the processor can be other general-purpose processors, DSPs, ASICs, FPGAs, or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0053] In one possible design, the memory is located outside the communication device, or inside the communication device, or the processor and memory are integrated together.

[0054] Fifthly, a chip system is provided, comprising a processor and potentially a communication interface for implementing the methods of the first or second aspect and their respective designs. Optionally, the chip system further comprises a memory. The memory is used to store computer programs (also referred to as code or instructions). The processor is used to retrieve and execute the computer programs from the memory, causing a device equipped with the chip system to perform the methods of the first or second aspect and their respective designs.

[0055] Sixthly, a communication device is provided, comprising an input / output interface and logic circuitry. The input / output interface is used for inputting and / or outputting information. The input / output interface may be an interface circuit, an output circuit, an input circuit, pins, or related circuitry, etc. The logic circuitry is used to execute the methods of the first or second aspect and their respective designs.

[0056] In one implementation, when the communication device is a wireless communication device, it can be a terminal device such as a mobile phone, or a network device such as a base station. The interface circuit can be a radio frequency processing chip in the wireless communication device, and the processing circuit can be a baseband processing chip in the wireless communication device.

[0057] In one implementation, when the communication device is a chip or chip system, the input circuit can be an input pin, the output circuit can be an output pin, and the logic circuit can be a transistor, gate circuit, flip-flop, or various other logic circuits. The input signal received by the input circuit can be received and input by, for example, but not limited to, a receiver, and the signal output by the output circuit can be, for example, but not limited to, output to a transmitter and transmitted by the transmitter. Furthermore, the input circuit and the output circuit can be the same circuit, which is used as both the input circuit and the output circuit at different times. This application does not limit the specific implementation of the input / output interface and the logic circuit.

[0058] The aforementioned communication device can be a terminal device or a component / part of a terminal device. For example, the communication device can be a communication module and / or computing module of a terminal device; or the communication device can be a circuit or chip in the terminal device responsible for communication functions (such as a modem chip, or a SoC chip or SIP chip containing a modem core); or the communication device is a circuit or chip in the terminal device responsible for communication and / or computing functions (such as a CPU, DSP, GPU, NPU, AI processor, ASIC, or FPGA); or the communication device is a logic node or logic module capable of implementing all or part of the functions of the terminal device.

[0059] The aforementioned communication device can be a network device or a component / part of a network device. For example, the communication device can be a communication module and / or computing module of a network device; or the communication device can be a circuit or chip responsible for communication functions in a network device (such as a modem chip, or a SoC chip or SIP chip containing a modem core); or the communication device is a circuit or chip responsible for communication and / or computing functions in a network device (such as a CPU, DSP, GPU, NPU, AI processor, ASIC, or FPGA); or the communication device is a logical node or logical module capable of implementing all or part of the functions of a network device.

[0060] In a seventh aspect, a communication system is provided, comprising a first device and a second device. The first device is used to implement the functions described in the first aspect, and the second device is used to implement the functions described in the second aspect. For example, the second device sends first information to the first device, the first information indicating a correlation between the performance of a first AI function and the performance of a second AI function; the first device acquires the performance of the first AI function and determines the performance of the second AI function based on the first information and the performance of the first AI function.

[0061] Eighthly, a computer-readable storage medium is provided, which stores a computer program or readable instructions, such that when a computer reads and executes the computer program or readable instructions, the computer performs the method described in any of the first or second aspects and any of their designs.

[0062] Ninth aspect, a computer program product is provided, which, when read and executed by the settlement price, causes a computer to perform the method described in the first or second aspect and any one thereof.

[0063] The beneficial effects of the third to ninth aspects mentioned above can be referenced to the beneficial effects of the first or second aspects and any of their designs. Attached Figure Description

[0064] Figures 1 and 2 are schematic diagrams of the architecture for deploying AI functions;

[0065] Figure 3 is a schematic diagram of the process by which the base station determines the applicability of AI functions on the UE side.

[0066] Figure 4 is a schematic diagram of the AI ​​function detection process;

[0067] Figure 5 is a schematic diagram of the architecture of a communication system;

[0068] Figures 6 and 7 are schematic flowcharts of the communication method provided in the embodiments of this application;

[0069] Figures 8 and 9 are schematic diagrams of the communication device provided in the embodiments of this application. Detailed Implementation

[0070] In the embodiments of this application, "transmission" includes "sending" and / or "receiving". "Sending" and "receiving" indicate the direction of signal transmission. "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.

[0071] For example, "sending information" can occur between devices or between logical modules (such as components, modules, chips, software modules, or hardware modules) within a device. For instance, "a network device sending information" can be understood as a network device sending information to another device (such as a terminal device), or it can be understood as logical module 1 in the network device sending information to logical module 2 in the network device.

[0072] Similarly, "receiving information" can occur between devices or between logical modules (such as components, modules, chips, software modules, or hardware modules) within a device. For example, "a network device receiving information" can be understood as a network device receiving information from another device (such as a terminal device), or it can be understood as logical module 1 in the network device receiving information from logical module 2 in the network device.

[0073] For example, phrases like "sending information to XX (e.g., a terminal device)" or the sending-related illustrations in the accompanying drawings can be understood as the destination of the information being XX, including sending information directly or indirectly to XX. Similarly, phrases like "receiving information from YY (e.g., a network device)," "receiving information from YY (e.g., a network device)," or "receiving information sent by YY (e.g., a network device)," or the receiving-related illustrations in the accompanying drawings, can be understood as the source of the information being YY, including receiving information directly or indirectly from YY. Furthermore, information may undergo necessary processing between the source and destination, such as format changes, but the destination can understand the valid information from the source. Similar expressions in this application can be interpreted similarly and will not be elaborated further here.

[0074] In the embodiments of this application, the number of nouns, unless otherwise specified, refers to "singular nouns or plural nouns," that is, "one or more." "At least one" means one or more, and "more than one" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, or B exists alone, where A / B can be singular or plural. The character " / " generally indicates that the related objects before and after are in an "or" relationship. For example, A / B means: A or B. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, and / or c means the following combination: a exists alone, b exists alone, c exists alone, a and b exist simultaneously, a and c exist simultaneously, b and c exist simultaneously, or a, b, and c exist simultaneously, where a, b, and c can be single or multiple.

[0075] In the embodiments of this application, "when," "if," and "if" all refer to the device taking corresponding actions under certain objective circumstances, not to a time limit, nor to requiring the device to perform a judgment action, nor implying any other limitations. Unless otherwise specified, "if" and "if" are interchangeable, and "when" is interchangeable with "in the case of," "if" / "if." Words such as "exemplary" or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or designs. The use of words such as "exemplary" or "for example" is intended to present related concepts in a specific manner. Furthermore, "of," "corresponding (relevant)," and "corresponding" can sometimes be used interchangeably; it should be noted that their intended meanings are consistent when their differences are not emphasized. In the embodiments of this application, "corresponding" can also be replaced with "corresponding," "associated," or "mapped." "A and B correspond" can be understood as "A and B have an associative / corresponding relationship." There are no restrictions on the specific implementation form of the correspondence / association relationship. For example, the correspondence relationship can be a mapping table, a function relationship, etc.

[0076] In embodiments of this application, "for indicating" can include both direct and indirect indication. For example, when describing a certain indication information as indicating information I, it can include whether the indication information directly indicates I or indirectly indicates I, but does not necessarily mean that the indication information carries I.

[0077] The terms "first," "second," "#1," "#2," and "#A" used in this application's embodiments are merely for descriptive convenience and to distinguish multiple objects. They are not intended to limit the size, content, order, timing, priority, or importance of the multiple objects. For example, "first AI function" and "second AI function" refer to two different AI functions, and do not indicate a difference in priority or importance between the two AI functions.

[0078] Before introducing the solutions provided in the embodiments of this application, the relevant terms / concepts involved in the embodiments of this application will be explained first.

[0079] (1) Access network equipment

[0080] Access network equipment refers to radio access network (R)AN equipment / RAN nodes. R)AN and RAN are interchangeable; for ease of description, RAN will be used as an example below. RAN can be a cellular system related to the 3rd generation partnership project (3GPP), such as a 5th generation (5G) mobile communication system / new radio (NR) communication system, or a future-oriented evolution system. RAN can also be an open RAN (O-RAN or ORAN), a cloud radio access network (CRAN), a virtualized RAN (vRAN), an artificial intelligence radio access network (AI RAN), a non-terrestrial network (NTN) (e.g., satellite communication systems), etc. RAN can also be a communication system that integrates two or more of the above systems.

[0081] In one possible scenario, a RAN node can be a base station, an evolved NodeB (eNodeB), an access point (AP), a transmission reception point (TRP), a next-generation nodeB (gNB), a satellite, or a base station in a future mobile communication system. A gNB is interchangeable with a next-generation radio access network (NG-RAN). RAN nodes can also be macro base stations, micro base stations, indoor stations, relay nodes, donor / host nodes, or radio controllers. RAN nodes can also be servers, wearable devices, vehicles, or in-vehicle equipment. For example, in vehicle-to-everything (V2X) technology, the RAN node can be a roadside unit (RSU). An AP can serve as the central hub of this communication system and can be a base station with a Wi-Fi chip, a router, a gateway, a repeater, a communication server, a switch, or a bridge. RAN nodes can also be satellites (or satellite base stations) or high altitude platform stations (HAPS), or base station equipment mounted on satellites / HAPS. RAN nodes can also be gateway stations (or ground stations, earth stations, signaling stations, gateways, or gateway stations).

[0082] RAN nodes, also known as RAN entities or access nodes, form part of the communication system and help terminal devices achieve wireless access. In future scenarios, RAN nodes may evolve into other forms; for example, RAN nodes may not be distinguished from core network equipment and may be collectively referred to as network equipment. Unless otherwise specified, in the embodiments of this application, access network equipment and network equipment can be used interchangeably.

[0083] In another possible scenario, the RAN node can be a module or unit that performs some of the functions of the base station; or multiple RAN nodes can cooperate to assist terminal equipment in achieving wireless access, with different RAN nodes performing some of the functions of the base station. For example, the RAN node can be a CU, DU, or RU. The function of the CU can be implemented by a single entity or by different entities. For example, the function of the CU can be further divided, that is, the control plane and the user plane can be separated and implemented by different entities, namely the control plane CU entity (i.e., CU-control plane (CP) entity) and the user plane CU entity (i.e., CU-user plane (UP) entity). The CU-CP entity and the CU-UP entity can be coupled with the DU to jointly complete the function of the RAN node. The CU and DU can be set up separately or included in the same network element, such as in the baseband unit (BBU). Any of the units among the CU (or CU-CP, CU-UP), DU, and RU in this application can be implemented by software modules, hardware modules, or a combination of software modules and hardware modules.

[0084] In different systems, CU (or CU-CP and CU-UP), DU, or RU may have different names, but those skilled in the art will understand their meaning. For example, in an ORAN system, CU can also be called O-CU (open CU), DU can also be called O-DU, CU-CP can also be called O-CU-CP, CU-UP can also be called O-CU-UP, and RU can also be called O-RU. For ease of description, this application uses CU, CU-CP, CU-UP, DU, and RU as examples.

[0085] The CU and DU can be configured according to the protocol layer functions of the wireless network they implement. For example, the CU can be configured to implement the functions of the Packet Data Convergence Protocol (PDCP) layer and higher protocol layers (such as the Radio Resource Control (RRC) layer and / or the Service Data Adaptation Protocol (SDAP) layer). The CU connects to network nodes such as the core network through interfaces, which can be E2 interfaces. Optionally, the CU has some core network functions. The CU (such as the PDCP layer and higher layers) connects to the DU (such as the Radio Link Control (RLC) layer and lower layers) through interfaces. For example, the DU can be configured to implement the functions of protocol layers below the PDCP layer (such as the RLC, Medium Access Control (MAC) layer, and / or Physical (PHY) layer). For specific descriptions of the above protocol layers, please refer to the relevant 3GPP technical specifications or the technical specifications of other applicable communication protocols.

[0086] The above division of CU and DU processing functions according to the protocol layer is merely an example; other division methods are also possible, and this application does not impose any restrictions.

[0087] For example, in one design, the CU or DU can be further divided into processing functions with protocol layers. In one design, some functions of the RLC layer and the protocol layer functions above the RLC layer are located in the CU, while the remaining functions of the RLC layer and the protocol layer functions below the RLC layer are located in the DU. In another possible design, the DU and RU cooperate to implement the PHY layer functions, or it can be described as moving some of the PHY layer functions of the DU to the RU. A DU can be connected to one or more RUs. The functions of the DU and RU can be configured in various ways depending on the design. For example, the DU is configured to implement baseband functions, and the RU is configured to implement mid-RF functions. As another example, the DU is configured to implement higher-level functions in the PHY layer, and the RU is configured to implement lower-level functions in the PHY layer, or to implement both lower-level and RF functions.

[0088] In the embodiments of this application, all or part of the functions of the RAN node can also be implemented through software functions running on hardware, or through virtualization functions instantiated on a platform (e.g., a cloud platform). The RAN node may also be equipped with communication modules, circuits, or chips that perform corresponding communication functions. The RAN node may also be configured with program instructions for performing corresponding communication functions and corresponding program instructions. The RAN node in the embodiments of this application may also be a logical node, logical module, or software function capable of implementing all or part of the access node functions, or a circuit or chip (such as a GPU, AI processor, NPU, ASIC, or FPGA) responsible for communication and / or computing functions in the access node.

[0089] A RAN can deploy one or more AI functions. As shown in Figure 1, when the RAN is an O-RAN, the O-RAN includes an intelligent controller. The intelligent controller can be a non-real-time RAN intelligent controller (RIC / non-RT RIC / NRT RIC) or a near-real-time RAN intelligent controller (RIC / near-RT RIC / nRT RIC).

[0090] Non-real-time RICs can be used to implement non-real-time intelligent management of RAN functions, enabling workflows including model training and model inference, and guiding applications / functions within the nRT RIC based on policies. For example, it can be used to train AI models and then use those models for inference. Non-real-time RICs can obtain network-side and / or terminal-side information from RAN nodes (e.g., CUs, CU-CPs, CU-UPs, DUs, compute nodes, and / or RUs) and / or terminals. This information can be used as training data or inference data, and the inference results can be delivered to RAN nodes and / or terminals. Inference results can be exchanged between CUs and DUs, and / or between DUs and RUs; for example, a non-real-time RIC delivers inference results to a DU, which then forwards them to an RU.

[0091] Near real-time RICs can be used to achieve near real-time intelligent management of the RAN. Through data collection and related operations on the E2 interface, near real-time control and optimization of O-RAN modules and resources can be achieved. For example, it can be used to train AI models and then use those models for inference. Non-real-time RICs can obtain network-side and / or terminal-side information from RAN nodes (e.g., CUs, CU-CPs, CU-UPs, DUs, compute nodes, and / or RUs) and / or terminals. This information can be used as training data or inference data, and the inference results can be delivered to RAN nodes and / or terminals. Inference results can be exchanged between CUs and DUs, and / or between DUs and RUs; for example, a non-real-time RIC delivers inference results to a DU, which then forwards them to an RU.

[0092] Near real-time RICs and non-real-time RICs can also be configured as separate network elements. Near real-time and non-real-time RICs can also be part of other devices; for example, near real-time RICs can be configured in RAN nodes (e.g., CU, DU, compute nodes), while non-real-time RICs can be configured in other network devices, such as core network devices or the cloud.

[0093] (2) Terminal equipment

[0094] In this application embodiment, anything capable of data communication with a base station can be considered a terminal device. Terminal devices are also called terminals, terminal equipment, user equipment (UE), mobile stations, or mobile terminals, etc. Terminal devices can be widely used in various scenarios, such as device-to-device (D2D), V2X communication, machine-type communication (MTC), Internet of Things (IoT), virtual reality (VR), augmented reality (AR), industrial control, autonomous driving, telemedicine, smart grids, smart furniture, smart offices, smart wearables, smart transportation, smart cities, etc. Terminal devices can be: mobile phones, computers, mobile internet devices (MID), wearable devices, VR devices, AR devices, stations (STA), robotic arms, cameras, robots, vehicles, drones, or smart home devices (such as televisions, air conditioners, robot vacuums, speakers), relays, customer premises equipment (CPE), etc. Among them, STA can be mobile phones, tablets, smart TVs, smart wearable devices, vehicle communication devices, routers, switches, etc., that support Wi-Fi communication. Terminal devices can also be communication modules, satellite phones or their components with satellite communication capabilities, or satellite communication terminals, such as very small aperture terminals (VSAT) (commonly referred to as VSAT terminals), portable stations, fixed stations, vehicle-mounted or airborne satellite communication terminals, etc. It should be understood that satellite communication terminals can serve as micro base stations to further provide data interfaces to accessed user equipment.

[0095] The embodiments of this application do not limit the specific technology or device form used in the terminal device. The terminal device typically includes a communication module, circuit, or chip that performs corresponding communication functions. Furthermore, it may also include modules, circuits, or chips (such as GPUs, AI processors, NPUs, or ASICs) that perform corresponding communication and / or computing functions. The terminal may also be configured with program instructions for performing corresponding communication and / or computing functions. In addition, when the terminal device is applied to V2X, it may also be referred to as a V2X device.

[0096] The various terminal devices described above, if located on a vehicle (e.g., placed / installed inside the vehicle), can all be considered in-vehicle terminal devices. In-vehicle terminal devices can be built into a vehicle's in-vehicle module, in-vehicle component, in-vehicle chip, or in-vehicle unit as one or more components or units. The vehicle can implement the methods of this application through the built-in in-vehicle module, in-vehicle component, in-vehicle chip, or in-vehicle unit. In-vehicle terminal devices can be vehicle equipment, in-vehicle modules, vehicles, in-vehicle units (on-board units, OBUs), remote sensing units (RSUs), in-vehicle infotainment systems (or in-vehicle transmission units) (telematics boxes, T-boxes), chips, or system-on-chips (SoCs), etc. These chips or SoCs can be installed in the vehicle, OBU, RSU, or T-box.

[0097] Figure 2 illustrates a communication system. This system includes access network equipment and terminal equipment. Figure 2 uses an example with two access network devices (or RAN nodes), both of which are CU-DU separated architectures (or the RAN is O-RAN). The CU and DU communicate via the FI interface, and the CU and DU communicate directly via the Xn interface. The terminal equipment can also deploy one or more AI models. The access network devices can also deploy one or more AI models. One or more AI models can be deployed on the CU and / or DU of the access network device.

[0098] (3) AI Function

[0099] In this application embodiment, an AI function refers to an AI feature / feature group enabled by one or more configurations (such as radio resource control (RRC) configurations), which can be determined by applicable conditions supported by the UE. For example, if CSI measurement resource configuration 1 and CSI reporting configuration 1 meet the applicable conditions for the UE to use AI for AI CSI feedback, then the AI ​​CSI reporting function based on CSI measurement resource configuration 1 and CSI reporting configuration 1 can be referred to as an AI function.

[0100] AI functions are implemented based on AI models, which can be deployed on AI modules, AI nodes, AI network elements, or AI entities. For ease of description, AI modules, AI nodes, AI network elements, or AI entities are collectively referred to as AI devices. AI devices can be independent devices, integrated into the same device to implement different functions, or they can be components in hardware devices, software functions running on dedicated hardware, or virtualized functions instantiated on a platform (e.g., a cloud platform). This application does not limit the specific form of AI devices. Alternatively, AI devices can communicate with other devices in a communication system, or AI devices can be circuits or chips in other devices, or AI devices can be logical nodes, logical modules, or software that can implement all or part of the functions of other devices. Other devices can be one or more of the following: access network devices, terminal devices, or core network devices (or network elements in the core network), etc. For example, the AI ​​model can be deployed in one or more of the following locations in the communication system shown in Figure 2: access network equipment (RAN node), terminal equipment, etc.; or the AI ​​device can also be deployed independently, for example, in a location other than any of the above-mentioned devices, such as in the host or cloud server of an over-the-top (OTT) system.

[0101] An AI model is a concrete implementation of AI functionality, representing the mapping relationship between the model's input and output. AI models can be neural networks, linear regression models, decision tree models, support vector machines (SVM), Bayesian networks, Q-learning models, or other machine learning models. In one example, the aforementioned neural network could be a deep neural network (DNN), a convolutional neural network (CNN), a recurrent neural network (RNN), or a generative adversarial network (GAN).

[0102] In this embodiment, the usage process involving the AI ​​function or AI model may include at least one of the following: data collection (collecting training data and / or inference data), data preprocessing, model training (or model learning), model information publishing (configuring model information), model transfer / transmission, model management (activation, deactivation, switching, rollback), model validation (validating the trained model), model inference (performing inference using the trained model), or publishing inference results. Inference can also be referred to as prediction. Optionally, the AI ​​model may include a machine learning (ML) model.

[0103] The AI ​​module's model can achieve different functions depending on the parameter configuration. The AI ​​module's model can be configured based on one or more of the following parameters: structural parameters (e.g., at least one of the following: number of neural network layers, neural network width, inter-layer connections, neuron weights, neuron activation function, or biases in the activation function), input parameters (e.g., the type and / or dimension of the input parameters), or output parameters (e.g., the type and / or dimension of the output parameters). The biases in the activation function can also be referred to as the neural network biases.

[0104] An AI device can have one or more models. A model can infer an output that includes one or more parameters. The learning, training, or inference processes of different models can be deployed on different nodes or devices, or they can be deployed on the same node or device.

[0105] (4) Typical applications of AI in air interface

[0106] AI applications in communications, particularly in air interface (AIA), have been extensively studied. For example, AI can be applied to modulation, coding, transmitters, receivers, multi-antenna technology, and positioning. Previous research projects have been initiated focusing on AIA air interface technology, examining its basic framework and several typical use cases / features (such as AI CSI feedback, AI beam management (BM), and AI positioning). The following sections will introduce the applications of AI in CSI feedback, beam management, and positioning.

[0107] (4-1) Application of AI in CSI Feedback

[0108] AI applications in CSI feedback include AI / ML-based CSI compression and AI / ML-based CSI prediction. AI / ML CSI compression uses AI / ML technology to compress and decompress downlink CSI information (measured by the terminal device). The terminal device transmits the compressed CSI information to the network device over the air interface, where it is then restored. Because AI / ML has a higher compression ratio and better CSI restoration capability compared to traditional compression algorithms, it can feed back more CSI information with less air interface overhead, enabling the network device to achieve more accurate downlink precoding. AI / ML CSI prediction uses AI / ML technology to predict downlink CSI for future times based on current / historical downlink CSI. Precoding based on the predicted CSI allows the predicted CSI to better match the channel state at the time the downlink data is scheduled, overcoming channel aging and achieving more accurate downlink precoding. The data used to train the AI ​​models for CSI compression and CSI prediction includes CSI data.

[0109] (4-2) Application of AI in Beam Management

[0110] The application of AI in beam management includes using AI / ML technology to predict the transmit and / or receive beams of network devices and / or terminal devices to obtain optimal beams. Compared to traditional beam management methods that require scanning a large number of candidate beams to obtain the optimal beam, AI / ML prediction of the better beam only requires scanning fewer beams, eliminating the need to scan all candidate beams. For example, in the spatial domain, fewer beams could be fewer sparse beams or wide beams, while a large number of candidate beams could be more dense beams or narrow beams. In the temporal domain, fewer beams could be beams from the current / historical time, while a large number of candidate beams could be all candidate beams from future time. For beam management, the data used to train the AI ​​model includes beam information, such as beam ID and the signal strength corresponding to the beam (e.g., reference signal receiving power (RSRP)).

[0111] (4-3) Application of AI in positioning

[0112] AI applications in positioning include using AI / ML technologies to infer intermediate parameters or location coordinates from measured channel information. Compared to traditional positioning methods, intermediate parameters or location coordinates obtained based on AI / ML technologies are more accurate. For positioning, the data used to train AI models includes channel information and / or location information. Optionally, channel information includes one or more of the following: power information, phase information, time delay information, distance information, velocity information, channel scattering information, line-of-sight (LOS) or non-line-of-sight (NLOS) information, etc.

[0113] (5) Unilateral and bilateral models

[0114] AI models over the air interface can be broadly categorized into two types: one-sided models and two-sided models. A two-sided model refers to a model where both the network side (e.g., base station) and the UE side have their own models. These two models need to be matched and work together to complete a function; therefore, they can be considered two parts of a single, unified model. Taking CSI compression as an example, the UE-side AI / ML model infers the original CSI to achieve compression and feeds back the inferred CSI (compressed CSI) to the base station via the air interface. Upon receiving the compressed CSI, the base station uses its own AI / ML model to infer the compressed CSI to achieve decompression, thus restoring the original CSI. For well-matched UE-side and base station-side AI / ML models, the restored CSI is relatively close to the original CSI; for poorly matched base station-side and UE-side AI / ML models, the restored CSI differs significantly from the original CSI.

[0115] A one-sided model refers to an AI model deployed only on the network side (e.g., base station) or the terminal side (e.g., UE). A one-sided model can independently complete the entire inference process for certain air interface characteristics. Taking beam management or positioning as an example, the entire inference process can be completed solely through an AI / ML model deployed on the base station or UE. Optionally, for network-side models, the UE may need to report measurement information to the base station as input data for model inference / training / monitoring; for terminal-side models, the UE may need to report inference results or monitoring results to the base station. In particular, for certain applications, the base station needs to instruct or configure corresponding resources (e.g., reference signal resources) for the UE to measure, in order to generate measurement information that needs to be reported to the base station or used for inference or detection in the UE-side model.

[0116] The AI ​​models involved in the various embodiments of this application can be either one-sided or two-sided models.

[0117] (6) Applicability of AI functions and management of the AI ​​function lifecycle

[0118] In real-world scenarios, AI functions are not always available. For example, whether an AI function can be used may vary depending on the environmental conditions, the capabilities or state of the device using the AI ​​function, or the configuration associated with the AI ​​function.

[0119] If, under specific conditions, the performance of an AI function meets specific requirements, or its performance metrics meet specific requirements, then the AI ​​function can be considered applicable under those specific conditions. Conversely, if, under specific conditions, the performance of an AI function does not meet specific requirements, or its performance metrics do not meet specific requirements, then the AI ​​function can be considered inapplicable under those specific conditions. For example, if an AI function is applicable in an indoor scenario, it is applicable indoors but not outdoors. Similarly, if an AI function is only applicable when the terminal device's battery level is greater than or equal to a certain threshold, then it is no longer applicable when the terminal device's battery level is less than the threshold.

[0120] For ease of description, the specific conditions under which AI functions are applicable are collectively referred to as applicable conditions. For example, applicable conditions may include one or more of the following: applicable environmental conditions, the capabilities of the device deploying the AI ​​function, the status of the device deploying the AI ​​function, or the configuration associated with the AI ​​function. Among these, applicable environmental conditions, the capabilities of the device deploying the AI ​​function, and the configuration associated with the AI ​​function can also be considered as different types of applicable conditions. For example, the capabilities of the device deploying the AI ​​function can be replaced with: the conditions that a device capable of performing AI functions must meet.

[0121] Environmental conditions can include the applicable scenarios for AI functions, the location of devices capable of performing AI functions, the speed of devices capable of performing AI functions, and the applicable channel conditions for AI functions, etc. For example, applicable scenarios for AI functions may include the applicable RSRP, signal-to-interference-plus-noise ratio (SINR), reference signal-received quality (RSRQ) and other signal quality requirements, channel characteristic requirements (such as the threshold range of power-delay profile (PDP) delay, the threshold range of channel impulse response (CIR) delay, and cell range requirements, etc. Environmental conditions also vary depending on the AI ​​function. For example, if an AI function is used for CSI feedback, the environmental conditions for that AI function may include channel conditions. As another example, if an AI function is used for positioning, the environmental conditions for that AI function may include the applicable scenarios (e.g., indoor or outdoor scenarios).

[0122] The capabilities of a device deploying AI functionality may include one or more of the following: the device's battery power, storage space, computing power, types of supported AI functions, supported AI estimation information, or the device's type. AI algorithms include at least one of the following: neural network algorithms, deep neural network algorithms, deep learning algorithms, federated learning algorithms, or other AI algorithms. For example, if the storage space required to run / implement a certain AI function exceeds a certain threshold, and the remaining or available storage space of the terminal device exceeds this threshold, then the AI ​​function is applicable to that terminal device. Conversely, if the remaining or available storage space of the terminal device is less than this threshold, then the AI ​​function is no longer applicable to that terminal device.

[0123] The status of a device that has deployed AI capabilities may include one or more of the following: the device’s remaining battery power (or available battery power), remaining storage space (or available storage space), remaining computing power (or available computing power), etc.

[0124] The configuration associated with AI functions can also be understood as running / implementing AI functions under this configuration. The configuration associated with AI functions can include some configurations for implementing AI functions, or configurations associated with AI functions. Taking the application of AI functions in beam management as an example, the configuration associated with AI functions can include one or more of the following: the reference signal pattern supported by the AI ​​function, the number of measurement reference signals, or the relationship between SetA and SetB. SetA is the wide beam set used in beam management, and SetB is the narrow beam set or a subset of the SetA beam set. Taking the application of AI functions in CSI feedback as an example, the configuration associated with AI functions can include the number of supported ports N, the number of supported measurement layers, etc.

[0125] The above applicable conditions are merely examples, and this application does not limit the names, types, number, or specific content of the applicable conditions. For example, for a two-sided model, the applicable condition for one AI function may also include its ability to match another AI function. Furthermore, one or more items referred to by a type of applicable condition can also be considered as an applicable condition. For example, the power consumption included in the capability of a device to deploy AI functions can be considered as an applicable condition.

[0126] Taking terminal devices as an example, they can determine the applicability of AI functions based on their applicable conditions. For instance, if a terminal device uses an AI function for positioning, and the power required to execute this function exceeds the device's remaining power, then the AI ​​function is no longer applicable. Similarly, if a terminal device uses an AI function for CSI feedback, and this function is configured for use in cell #1, then it is no longer applicable in cell #2. Furthermore, when a terminal device uses an AI function for time-domain prediction, if the device's speed falls within the failure speed threshold range corresponding to this AI function, it indicates that the AI ​​function cannot support longer measurement sample intervals, longer prediction sample intervals, or larger prediction windows, and is therefore no longer applicable. Additionally, a terminal device can determine the delay distribution characteristics through the channel's power-delay profile (PDP), thereby identifying its current scenario. When the delay distribution is concentrated in areas of low delay, it can be determined that the terminal device is in a LOS (Low-Installation-Outs) scenario or an NLOS (Low-Installation-Outs) scenario, thus determining whether the current AI function is applicable.

[0127] In some scenarios, the AI ​​functions on the UE side may be controlled by the network (NW) side, or in other words, the NW can manage the AI ​​functions on the UE side.

[0128] AI function management includes AI function lifecycle management (LCM). LCM manages the entire process of an AI model from start to finish, and can be understood as the steps / processes / strategies that an organization / individual should follow when using AI to obtain actual business value. A possible overall signaling flow for AI model LCM may include: AI-related capability interaction, data collection for AI model training, AI model training, AI model operation (including model registration, model configuration, etc.), data collection for AI model inference, AI model inference, AI model performance feedback and / or monitoring, and AI model switching / updating / deactivation / fallback to traditional methods. For ease of description, LCM can be referred to as AI function management in this application embodiment; the two terms are interchangeable. AI function management includes one or more of the following: activation, deactivation, updating, retraining, or switching of AI functions. In the process of using AI to implement the above use cases (such as CSI feedback), only the concept of the logical model can be presented externally. AI functions are usually implemented through AI models. The granularity of AI function management can be divided into two granularities: AI function and AI model. The granularity of AI function is coarser than that of AI model. When using AI, you can base it on the granularity of AI functions or the granularity of AI models.

[0129] The management of AI functions is based on at least one or more of the following aspects: the applicability of the AI ​​function, the performance of the AI ​​function, or performance metrics of the AI ​​function. For example, if the NW manages an AI function based on its applicability, it needs to understand whether the AI ​​function is applicable to the UE. If applicable, the AI ​​function can be activated; if not, it can be deactivated. Therefore, the UE can report relevant information to the NW, enabling the NW to manage the AI ​​function based on that information.

[0130] For applicable conditions unknown to the NW (such as the UE's remaining battery power or storage space), the NW can request the UE to report the applicable conditions of the AI ​​function, or the UE can proactively report the applicable conditions of the AI ​​function to the NW, and the NW can determine the applicability of the AI ​​function.

[0131] For example, please refer to Figure 3, which illustrates the process by which a base station determines the suitability of AI functions on the UE side. Figure 3 uses the application of AI functions in beam management as an example, and this process may include steps 301 to 302.

[0132] Step 301: The base station sends the model inference configuration for the AI ​​function to the UE.

[0133] Model inference configurations can include configurations associated with AI functions (e.g., SetA / SetB), providing structured input data to the AI ​​model to support subsequent analysis and prediction. Model inference configurations can be hosted in one or more fields used for AL / ML, and a field can include one or more information elements.

[0134] Step 302: The UE sends an AI function suitability report to the base station.

[0135] This suitability report indicates the suitability of AI functions under current conditions (e.g., configured SetA / SetB, channel state). The suitability report may be carried within one or more fields included in the suitability reporting fields used for AL / ML, and a field may include one or more information elements.

[0136] (7) Performance testing of AI functions

[0137] AI functions are implemented through AI models, which can be closely integrated with specific application scenarios and conditions. For example, AI models can be trained specifically for these scenarios or conditions. When using an AI model to implement a function, the performance of that function is closely related to the performance of the AI ​​model. To ensure optimal system performance, the performance of the AI ​​model is usually tested. An AI model has two phases: an inactive phase and an active phase. For an inactive AI model, its performance is primarily characterized by its applicability. In other words, testing the performance of an inactive AI model mainly involves testing its applicability. For an active AI model, its performance is characterized by its output results. For example, a higher accuracy in the output results indicates higher performance, while lower accuracy indicates lower performance.

[0138] For example, please refer to Figure 4, which illustrates the detection process for AI functions. Figure 4 uses the application of AI functions in positioning as an example, and the detection process for this AI function may include steps 401 to 403.

[0139] Step 401: The base station sends a positioning reference signal configuration to the UE.

[0140] The configuration of the positioning reference signal may include the resource configuration of the positioning reference signal and whether the positioning reference signal is periodic.

[0141] Step 402: The base station sends a positioning reference signal to the UE.

[0142] The base station sends a positioning reference signal to the UE, and the UE measures the positioning reference signal to determine its location. In this embodiment, the UE measures the positioning reference signal based on an AI model to achieve positioning functionality based on AI / ML.

[0143] Step 403: The UE sends the measurement results of the AI ​​model to the base station.

[0144] The base station can determine the performance of the AI ​​model based on the measurement results. For example, if the positioning accuracy indicated by the measurement results is greater than or equal to a preset threshold, the AI ​​model's performance is good; if the positioning accuracy indicated by the measurement results is less than the preset threshold, the AI ​​model's performance is poor. Similarly, if the positioning latency indicated by the measurement results is greater than or equal to a preset threshold, the AI ​​model's performance is poor; if the positioning accuracy indicated by the measurement results is less than the preset threshold, the AI ​​model's performance is good.

[0145] The base station determines the performance of the AI ​​function based on the applicability report reported by the UE. If the applicability report indicates that the AI ​​function is no longer applicable, then the performance of the AI ​​function is low.

[0146] Figure 4 uses an activated AI model as an example, where the UE reports the measurement results of the AI ​​model. It should be understood that for an inactive AI model, the base station can configure at least one set of parameters for the UE, and the UE reports the applicability of at least one set of parameters to the base station. For an applicable set of parameters, the base station can enable that set of parameters to activate the AI ​​model. For an activated AI model, the performance of the AI ​​model can be tested using the process shown in Figure 4.

[0147] As shown in Figure 4, detecting an AI model requires at least three signaling communications. In the future, there will be more AI functions / models. If each AI function / model independently tests its performance, the signaling will become increasingly large, leading to higher resource overhead. Using the example in Figure 4, detecting the performance of M AI models requires 3 × M signaling communications. The larger M is, the greater the signaling overhead. Similarly, in Figure 4, the base station needs to send positioning reference signals. Detecting the performance of M AI models requires sending M positioning reference signals. The larger M is, the greater the overhead of the positioning reference signals.

[0148] Therefore, the present application provides a solution based on embodiments of this application. This application considers that the performance of different AI functions may be correlated; thus, the performance of one AI function (e.g., the first AI function) can be used to infer the performance of another AI function (e.g., the second AI function). With the solution of this application, only the performance of the first AI function needs to be detected to know the performance of the second AI function, eliminating the need to separately detect the performance of the first and second AI functions. This reduces the number of AI functions to be detected, thereby saving signaling overhead for detecting AI function performance in the system, and even saving resource overhead for detecting AI function performance in the system.

[0149] The technical solutions provided in the embodiments of this application can be applied to various communication systems, such as long-term evolution (LTE) communication systems, 5G mobile communication systems / NR communication systems, or future mobile communication systems, or other similar communication systems. Other similar communication systems may include V2X systems, IoT systems, NTN, or wireless local area networks (WLANs), etc. WLANs can be WLANs using any of the protocols in the Institute of Electrical and Electronics Engineers (IEEE) 802.11 series. The methods provided in the embodiments of this application can also be applied to scenarios where NTN and terrestrial networks (TN) are integrated. The NTN system can be an NTN system integrated with 4G, 5G, and any future generation of communication systems, such as NR NTN, IoT NTN, etc. The NTN communication system can be, for example, a satellite communication system, and may also include drones, HAPS, and other airborne access network equipment; this application does not limit this. Furthermore, the technical solutions provided in this application can be applied to D2D scenarios, such as NR-D2D scenarios, or to V2X scenarios, such as NR-V2X scenarios. For example, they can be used in fields such as factory manufacturing, smart homes, intelligent driving, assisted driving, or intelligent connected vehicles.

[0150] Figure 5 illustrates a communication network architecture applicable to an embodiment of this application. The communication network may include a first device and a second device, which are capable of communicating with each other. In some scenarios, the first device may be a terminal device, and the second device may be a network device, as illustrated in Figure 5. In other scenarios, both the first and second devices may be terminal devices, such as a first terminal device and a second terminal device. The first device and / or the second device may deploy one or more AI models to achieve certain functions. The AI ​​model deployed by the first device may include a one-sided model and / or a two-sided model. The AI ​​model deployed by the second device may also include a one-sided model and / or a two-sided model. For example, the first device is a terminal device, and the second device is a network device. The terminal device deploys a first AI model, which can be used to perform channel estimation to determine the channel information between the terminal device and the network device. Optionally, the network device deploys a second AI model, which works in conjunction with the first AI model to perform channel estimation. As another example, both the first and second devices are terminal devices. The first terminal device may deploy a first AI model, which can be used for positioning. The first terminal device can then notify the second terminal device of the positioning results based on the first AI model. The second terminal device can deploy a second model, which can be used to schedule multiple terminal devices based on their locations to reduce congestion.

[0151] In the following description, the communication method provided in the embodiments of this application is applied to the architecture shown in Figure 5 as an example. The network architecture and application scenario shown in Figure 5 are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided in the embodiments of this application. As those skilled in the art will know, with the evolution of network architecture and the emergence of new application scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems. For example, Figure 5 may include more or fewer network devices and / or terminal devices. As another example, the communication system may also include other devices, such as core network devices, which are not shown in Figure 5. When applying the technical solutions of the embodiments of this application to other communication systems, the devices, components, modules, etc. in the embodiments can be replaced with corresponding devices, components, modules in other communication systems without limitation.

[0152] The solutions provided in this application are described below with reference to specific embodiments.

[0153] In the following description, the methods in the embodiments of this application are implemented using terminal devices and network devices as examples. The network device described in the various embodiments of this application can be the access network device shown in FIG5, and the terminal device described in the various embodiments of this application can be the terminal device shown in FIG5. In addition, the network device in the embodiments of this application can be a core network device.

[0154] The steps performed by the network device can also be performed by hardware and / or software components applicable to the network device. For example, the steps performed by the network device can also be performed by circuits or chips in the network device responsible for communication functions (such as modem chips (also known as baseband chips), or SoC chips or SIP chips containing modem cores), chip systems, or processors; they can also be performed by circuits or chips in the network device responsible for communication and / or computing functions (such as CPUs, GPUs, NPUs, AI processors, ASICs, or FPGAs); or they can be implemented by logical nodes (or logical modules) and / or software that implement all or part of the functions of the network device.

[0155] The steps performed by the terminal device can also be performed by hardware and / or software components applicable to the terminal device. For example, the steps performed by the terminal device can also be performed by circuits or chips in the terminal device responsible for communication functions (such as modem chips (also known as baseband chips), or SoC chips or SIP chips containing modem cores), chip systems, or processors; they can also be performed by circuits or chips in the terminal device responsible for communication and / or computing functions (such as CPUs, GPUs, NPUs, AI processors, ASICs, or FPGAs); or they can be implemented by logical nodes (or logical modules) and / or software that implement all or part of the functions of the terminal device.

[0156] In this embodiment of the application, the processing performed by a single execution entity can also be divided into multiple execution entities, which can be logically and / or physically separated. For example, the processing performed by a network device can be divided into at least one execution entity among CU, DU, RU, etc.

[0157] In the various embodiments of this application, "AI model," "AI function," "ML model," "ML function," "AI feature," "AI algorithm," "ML feature," and "ML algorithm" can be used interchangeably. The AI ​​model in the various embodiments of this application can be a one-sided model or a two-sided model. The device that "deploys" the AI ​​model can refer to the device that runs the AI ​​model; "deployment" can also be understood as "running," "using," or "storing." For example, in Figure 5, if the terminal device is the device that deploys the AI ​​model, then the terminal device can run the AI ​​model, perform inference processes through the AI ​​model, and obtain the inference results output by the AI ​​model.

[0158] In various embodiments of this application, the AI ​​model or AI function can be trained by a terminal device; alternatively, the AI ​​model or AI function can be trained by a network device, which, after training, can send the parameters of the AI ​​model or AI function to the terminal device, allowing the terminal device to deploy the AI ​​model or AI function; alternatively, the AI ​​model or AI function can be trained by a third-party device, which, after training, can send the parameters of the AI ​​model or AI function to the terminal device, allowing the terminal device to deploy the AI ​​model or AI function; or, after training, the third-party device can send the parameters of the AI ​​model or AI function to a network device, which then sends the parameters to the terminal device, allowing the terminal device to deploy the AI ​​model or AI function. The third-party device may be, for example, a network management (OAM) server, an over-the-top (OTT) server, or other devices.

[0159] In the various embodiments of this application, the performance detection of the AI ​​function refers to detecting, evaluating, and testing the performance of the AI ​​function. The performance detection of the AI ​​function can also be replaced by: performance evaluation or testing of the AI ​​function, etc. The applicable conditions of the AI ​​function can be replaced by the effective conditions of the AI ​​function; the applicability of the AI ​​function can also be understood as the AI ​​function being effective, and correspondingly, the failure of the AI ​​function can be understood as the AI ​​function being inapplicable. Unless otherwise specified, in the accompanying drawings corresponding to the various embodiments of this application, all steps represented by dashed lines are optional steps.

[0160] Example 1

[0161] Please refer to Figure 6, which is a flowchart illustrating the first communication method provided in this embodiment. Figure 6 describes the method from the perspective of interaction between network devices and terminal devices. As shown in Figure 6, the communication method includes the following steps.

[0162] S600: The network device sends a first request to the terminal device, and the terminal device receives the first request from the network device accordingly.

[0163] A terminal device may deploy one or more AI models, or some or all of the terminal device's functions may be implemented through AI models. These one or more AI models include unilateral and / or bilateral models. In this embodiment, the network device may detect the performance of one or more AI models on the terminal device. Since AI models are used to implement AI functions, detecting the performance of AI models can also be referred to as detecting the performance of AI functions.

[0164] The first request can be used to request AI functions that are performance-related, or the first request can be used to request AI functions that are performance-related. AI functions that are performance-related will be described in detail in S601 below, and will not be discussed here.

[0165] Optionally, the network device can request other AI functions associated with the performance of a certain AI function. For example, if the network device requests an AI function associated with the performance of a second AI function, the first request may include the ID of the second AI function. In this case, the terminal device only reports information about the AI ​​functions associated with the performance of the second AI function, saving signaling resource overhead.

[0166] Optionally, when the network device needs to test the performance of one or more AI functions, it sends a first request to the terminal device. The specific name of the first request is not limited; for example, it can be called a query request, a query request for performance-related AI functions, or a query request for applicability-related AI functions. The signaling carrying the first request is not limited; for example, the first request can be carried in an RRC message, a MAC control element (CE), or downlink control information (DCI). Furthermore, S600 is not a mandatory step and is therefore illustrated with dashed lines in Figure 6.

[0167] S601, The terminal device sends first information to the network device, and the network device receives the first information from the terminal device accordingly.

[0168] The first information may include information used to / assist in determining the performance of AI functions. Network devices can determine the performance of one or more AI functions based on the first information. There is no limitation on the specific name of the first information; for example, it may be called AI function performance detection information (or simply AI performance detection information), AI function performance evaluation information (or simply AI performance evaluation information), or AI performance assistance information, etc. Optionally, the first information may serve as a capability information of the terminal device; therefore, it may also be called UE capability information. There is no limitation on the signaling carrying the first information; for example, the first information may be carried in RRC messages, uplink control information (UCI), UE assistance information (UAI), or UE capability information.

[0169] Optionally, the terminal device receives a first request and, in response to the first request, sends first information to the network device. Alternatively, the terminal device decides to send the first information to the network device on its own. For example, the terminal device periodically sends the first information to the network device, or the terminal device sends the first information to the network device when information assisting in determining the AI ​​function has been updated / changed. The execution order of S601 and S600 is not limited. For example, after the terminal device sends the first information to the network device, the network device may also send a first request to the terminal device to obtain the latest first information or the latest performance-related AI function.

[0170] The first information can be used to indicate that the performance of multiple AI functions is related. For example, if the multiple AI functions include a first AI function and a second AI function, then the first information can be used to indicate that the performance of the first AI function is related to the performance of the second AI function.

[0171] In the embodiments of this application, if two AI functions meet specific conditions, then the performance of the two AI functions is related. For example, if two AI functions meet one or more of the following conditions (1) to (3), then the performance of the two AI functions is related.

[0172] (1) The performance of one AI function is better than or equal to that of the other AI function.

[0173] The performance of one AI function being better than or equal to that of another AI function includes situations where the performance of one AI function depends on / depends on another AI function.

[0174] For example, if the first AI function and the second AI function have a master-slave relationship, then the performance of the first AI function is always better than or equal to that of the second AI function. For instance, the first AI function is implemented based on a CSI estimation model, and the second AI function is implemented based on a CSI prediction model. The CSI estimation model is used to estimate the current CSI, and the CSI prediction model is used to predict the future CSI. If the performance of the CSI estimation model declines, then the performance of the CSI prediction model will also be poor. Therefore, it can be considered that the performance of the CSI estimation model is better than or equal to that of the CSI prediction model.

[0175] For example, if the first AI function fails to meet the first performance metric, it causes the second AI function to fail to meet the second performance metric. Conversely, if the first AI function meets the first performance metric, it causes the second AI function to meet the second performance metric. For instance, both the first and second AI functions are used for positioning. The first AI function is implemented based on a first AI model, and the second AI function is implemented based on a second AI model. The first AI model is a lightweight model, and the second AI model is a high-precision model. If the performance factor is positioning latency, the performance of the first AI model may always be better than or equal to the performance of the second AI model, or the performance of the first AI model reaching the first performance metric will cause the performance of the second AI model to always reach the second performance metric. If the performance factor is positioning accuracy, the performance of the second AI model may always be better than or equal to the performance of the first AI model, or the performance of the second AI model reaching the first performance metric will cause the performance of the first AI model to always reach the second performance metric.

[0176] (2) The applicable conditions of the two AI functions are related.

[0177] In Example 1, the related applicability conditions of two AI functions include: the applicability conditions of the same performance of the two AI functions are related, or the applicability conditions of different performances of the two AI functions are related. Taking the two AI functions as a first AI function and a second AI function as an example, the related applicability conditions of the two AI functions include: the applicability conditions of a first performance of the first AI function and the applicability conditions of a second performance of the second AI function are related. Here, the first performance and the second performance can be the same performance, or the first performance and the second performance can be different performances.

[0178] The terminal device can determine whether the applicable conditions of two AI functions are related based on one or more of the following: applicable environmental conditions, the capability of the device deploying the AI ​​function, the status of the device deploying the AI ​​function, or the configuration associated with the AI ​​function. For an explanation of the applicable environmental conditions, the capability of the device deploying the AI ​​function, the status of the device deploying the AI ​​function, or the configuration associated with the AI ​​function, please refer to the relevant content in the aforementioned terminology explanation (6), which will not be repeated here. The following uses specific examples to illustrate how the terminal device determines whether the applicable conditions of two AI functions are related.

[0179] (1-1) Determine whether the applicable conditions of the two AI functions are related based on the applicable environmental conditions of the two AI functions.

[0180] For example, both the first AI function and the second AI function are used for positioning. The first and second AI functions are applicable to the same environmental conditions, and the devices deploying them have the same capabilities. Both performance metrics are positioning accuracy. If the positioning accuracy of the first AI function is always greater than or equal to the positioning accuracy of the second AI function, then the applicable conditions of the first and second AI functions may be related. Conversely, if the positioning accuracy of the first AI function is not always greater than or equal to the positioning accuracy of the second AI function, then the applicable conditions of the first and second AI functions are not related. For example, the complexity of the first AI function may be lower than that of the second AI function, resulting in the positioning accuracy of the first AI function not always being greater than or equal to that of the second AI function.

[0181] For example, a first AI function is used for CSI compression, and a second AI function is used for CSI decompression. The devices using both functions have the same capabilities, and the configurations associated with both functions are identical. The first performance characteristic is compression, and the second is decompression. If the channel conditions applicable to both functions are the same, and the compression performance of the first AI function is good, the decompression performance of the second AI function will also always be good. In this case, the applicable conditions of the first and second AI functions may be related. Conversely, if the channel conditions applicable to both functions are the same, but the compression performance of the first AI function is good, the decompression performance of the second AI function is not always good. In this case, the applicable conditions of the first and second AI functions are not related.

[0182] (1-2) Determine whether the applicable conditions of the two AI functions are related based on the capabilities of the devices that deploy AI functions.

[0183] For example, both the first AI function and the second AI function are used for beam management. The configuration associated with the first AI function and the configuration associated with the second AI function are the same. The environmental conditions applicable to the first AI function and the second AI function are the same. The storage space required for the AI ​​model corresponding to running the first AI function is greater than a first threshold, and the storage space required for the AI ​​model corresponding to running the second AI function is greater than a second threshold. When the first threshold is less than the second threshold, if the first AI function is applicable to the first terminal device, then the second AI function is also applicable to the first terminal device, and the applicable conditions of the first AI function and the second AI function may be related; if the first AI function is not applicable to the first terminal device, then the second AI function is also not applicable to the first terminal device, and the applicable conditions of the first AI function and the second AI function are not related.

[0184] (1-3) Determine whether the applicable conditions of the two AI functions are related based on the status of the devices that have deployed AI functions.

[0185] For example, both the first AI function and the second AI function are used for beam management. The configuration associated with the first AI function and the configuration associated with the second AI function are the same. The environmental conditions applicable to the first AI function and the second AI function are the same. The power required for the AI ​​model corresponding to the first AI function is greater than a first threshold, and the power required for the AI ​​model corresponding to the second AI function is greater than a second threshold. When the first threshold is less than the second threshold, if the first AI function is applicable to the first terminal device, then the second AI function is also applicable to the first terminal device, and the applicable conditions of the first AI function and the second AI function may be related; if the first AI function is not applicable to the first terminal device, then the second AI function is also not applicable to the first terminal device, and the applicable conditions of the first AI function and the second AI function are not related.

[0186] (1-4) Determine whether the applicable conditions of the two AI functions are related based on the configuration associated with the AI ​​functions.

[0187] For example, both the first AI function and the second AI function are located in the positioning field, and the environmental conditions applicable to both are the same. If the configuration associated with the first AI function indicates that it supports periodic positioning reference signals, and the configuration associated with the second AI function indicates that it supports periodic positioning reference signals, then the applicable conditions of the first and second AI functions may be related. If the configuration associated with the first AI function indicates that it supports periodic positioning reference signals, and the configuration associated with the second AI function indicates that it supports non-periodic positioning reference signals, then the applicable conditions of the first and second AI functions are not related.

[0188] For example, both the first AI function and the second AI function are located in the CSI feedback. The configuration associated with the first AI function includes the number of ports measured (N1) and the number of layers measured (M1), while the configuration associated with the second AI function includes the number of ports measured (N2) and the number of layers measured (M2). If N1 is greater than N2 and M1 is greater than M2, then the applicable conditions of the first AI function and the second AI function may be related; if N1 is less than N2 and M1 is less than M2, then the applicable conditions of the first AI function and the second AI function are not related.

[0189] In Example 2, the applicable conditions of two AI functions are related in the following ways: the applicable conditions of the same or different performance indicators of the same performance of the two AI functions are related; or, the applicable conditions of the same or different performance indicators of different performance of the two AI functions are related. The applicable conditions of a performance indicator refer to the conditions under which that performance indicator is met.

[0190] Taking two AI functions, namely, the first AI function and the second AI function, as an example, the applicable conditions of the two AI functions are related, including the applicable conditions of the first performance index of the first AI function and the applicable conditions of the second performance index of the second AI function. Specifically, the first performance and the second performance can be the same performance, and the first performance index and the second performance index can be the same or different performance indices. Alternatively, the first performance and the second performance can be different performances, and the first performance index and the second performance index can be the same or different performance indices.

[0191] For example, both the first AI function and the second AI function are used for assisted localization, and both the first performance and the second performance are the recognition rate of LOS (Location of Observation). If, under the same conditions (such as environmental conditions and configuration associated with the AI ​​functions), the recognition rate of LOS for the first AI function is greater than the first threshold, and the recognition rate of LOS for the second AI function is always greater than the second threshold, then the conditions for the application of the first AI function and the second AI function may be related. Conversely, if the recognition rate of LOS for the first AI function is greater than the first threshold, and the recognition rate of LOS for the second AI function is not always greater than the second threshold, then the conditions for the application of the first AI function and the second AI function are not related.

[0192] For example, both the first AI function and the second AI function are used to assist in positioning. The first performance characteristic is positioning accuracy, and the second performance characteristic is the LOS recognition rate. If, under the same conditions (such as environmental conditions and configuration associated with the AI ​​functions), the positioning accuracy of the first AI function is greater than a first threshold, and the LOS recognition rate of the second AI function is always greater than a second threshold, then the conditions for the application of the first AI function and the second AI function may be related. Conversely, if the positioning accuracy of the first AI function is greater than the first threshold, and the LOS recognition rate of the second AI function is not always greater than the second threshold, then the conditions for the application of the first AI function and the second AI function are not related.

[0193] The implementation methods of AI functions with performance correlation, as illustrated below, include various approaches. In these examples, the AI ​​function ID only needs to be usable to identify the AI ​​function. For instance, the AI ​​function ID can be the ID of the AI ​​model corresponding to that AI function. For example, if the ID of AI model #1 is #1, the ID of the AI ​​function corresponding to AI model #1 can be "#1". Alternatively, the AI ​​function ID can be the ID of the configuration corresponding to that AI function. For example, if the network device is configured with configuration #A, which is used for / corresponds to AI function #1, then the ID of AI function #1 can be #A. Or, if the network device is configured with parameter set #A, which is used for / corresponds to AI function #1, then the ID of AI function #1 can be #A. In this paper, the AI ​​function ID can be replaced with the configuration ID, the ID of the configuration corresponding to the AI ​​function, the ID of the parameter set, or the ID of the parameter set corresponding to the AI ​​function. In this paper, the ID set includes one or more IDs, and the parameter set includes one or more parameters. Furthermore, there are no restrictions on the specific implementation form of the ID. For example, an ID can be a number / index assigned to an AI function, its corresponding configuration, or the set of parameters for that AI function. Alternatively, an ID can be a bit sequence associated with an AI function, its corresponding configuration, or the set of parameters for that AI function. Furthermore, in this paper, performance information may include one or more of the following: performance type information (such as the performance ID) or performance metric information. One AI function corresponds to one or more performance metrics, and one performance metric is associated with one or more performance metrics.

[0194] (2-1) The first information includes at least one set of IDs.

[0195] The IDs of AI functions with performance correlation form a set. For example, AI functions #1, #2, and #3 are performance-correlated, and their corresponding ID set is (#1, #2, #3); AI functions #4, #5, and #6 are performance-correlated, and their corresponding ID set is (#4, #5, #6). The first information includes at least one ID set of {(#1, #2, #3), (#4, #5, #6)}. For another example, configuration #A corresponds to AI function #1, configuration #B corresponds to AI function #2, configuration #C corresponds to AI function #3, configuration #D corresponds to AI function #4, and configuration #E corresponds to AI function #5. Assuming that AI functions #1, #2, and #3 are performance-correlated, and AI functions #4 and #5 are performance-correlated, then at least one ID set could be {(#A, #B, #C), (#D, #E)}.

[0196] Optionally, the first information may also include performance information associated with one or more ID sets. The performance information associated with an ID set may be performance information associated with some or all of the IDs in that ID set. For example, at least one ID set includes a first ID set, and the first information may also include performance information associated with some of the IDs in the first ID set, or the first information may also include performance information associated with all the IDs in the first ID set. The same ID set may correspond to one or more performance metrics, and different ID sets may correspond to different performance metrics.

[0197] For example, assuming that AI function #1, AI function #2 and AI function #3 are related in performance #1, and AI function #4, AI function #5 and AI function #6 are related in performance #2, then the first information can be {performance #1: (#1, #2, #3), performance #2: (#4, #5, #6)}.

[0198] For example, AI function #1, AI function #2 and AI function #3 are related in performance #1, and AI function #4, AI function #5 and AI function #6 are related in performance 1# and performance #2. Then the first information can be {performance #1: (#1, #2, #3) and {performance #1, performance #2}: (#4, #5, #6)}.

[0199] For example, AI functions #1, #2, and #3 are related in terms of performance #1 and performance #3, and AI functions #4, #5, and #6 are related in terms of performance #1 and performance #2. Then the first information could be {{performance #1, performance #3}: (#1, #2, #3), {performance #1, performance #2}(#4, #5, #6)}.

[0200] Optionally, the first information also includes at least one performance metric associated with one or more performance metrics. The one or more performance metrics are some or all of the at least one performance metric within the first information. A performance metric can be associated with one or more performance metrics. In this case, different sets of IDs may correspond to the same or different performance metrics.

[0201] For example, assuming AI functions #1, #2, and #3 are associated when performance #1 reaches performance metric #1, and AI functions #4, #5, and #6 are associated when performance #1 reaches performance metric #2, then the first information could be {performance #1 : {performance metric #1} : (#1, #2, #3), performance #1 : {performance metric #2} : (#4, #5, #6)}. As another example, assuming AI functions #1, #2, and #3 are associated when performance #1 reaches performance metric #1, and AI functions #4, #5, and #6 are associated when performance #2 reaches performance metric #1, then the first information could be {performance #1 : {performance metric #1} : (#1, #2, #3), performance #2 : {performance metric #1} : (#4, #5, #6)}.

[0202] For example, assuming AI functions #1, #2, and #3 are associated when performance #1 reaches performance metrics #1 and #2, and AI functions #4, #5, and #6 are associated when performance #1 reaches performance metric #3, then the first information could be {performance #1: {performance metrics #1, performance metrics #2}: (#1, #2, #3), performance #1: {performance metrics #3}: (#4, #5, #6)}. As another example, assuming AI functions #1, #2, and #3 are associated when performance #1 reaches performance metric #1, and AI functions #4, #5, and #6 are associated when performance #2 reaches performance metric #2, then the first information could be {performance #1: {performance metrics #1}: (#1, #2, #3), performance #2: {performance metric #2}: (#4, #5, #6)}.

[0203] (2-2) The first information includes at least one AI function ID and at least one set of IDs. Wherein, one set of IDs corresponds to one ID, including the IDs of AI functions associated with the performance of the AI ​​function indicated by the corresponding ID.

[0204] For example, AI function #1 is associated with the performance of AI functions #2 and #3 respectively, but the performance of AI function #2 is not associated with the performance of AI function #3. AI function #4 is associated with the performance of AI functions #5 and #6 respectively, but the performance of AI function #5 is not associated with the performance of AI function #6. The ID of at least one AI function includes #1 and #4, and at least one set of IDs can be {(#2, #3), (#5, #6)}, and the first information can be {#1:(#2, #3), #4:(#5, #6)}.

[0205] For example, AI function #1 is associated with the performance of AI functions #2 and #3 respectively, but the performance of AI function #2 is not associated with the performance of AI function #3. AI function #2 is associated with the performance of AI functions #5 and #6 respectively, but the performance of AI function #5 is not associated with the performance of AI function #6. The ID of at least one AI function includes #1 and #2, and at least one set of IDs can be {(#2, #3), (#5, #6)}, and the first information can be {#1:(#2, #3), #2:(#5, #6)}.

[0206] Optionally, the first information may also include performance information associated with one or more IDs, and / or performance information associated with one or more sets of IDs. Performance information associated with a set of IDs may be performance information associated with some or all of the IDs in that set. For example, if at least one ID includes a first ID, at least one ID set includes a first ID set, and the first ID and the first ID set correspond, the first information may also include performance information associated with the first ID. Alternatively, the first information may also include performance information associated with some or all of the IDs in the first ID set. Or, the first information may also include performance information associated with both the first ID and some or all of the IDs in the first ID set. The same ID set may correspond to one or more performance metrics, and different ID sets may correspond to different performance metrics. One performance metric may be associated with one or more performance indicators.

[0207] The IDs using at least one of the above AI functions include #1 and #4. For example, at least one set of IDs can be {(#2, #3), (#5, #6)}. The first information can be {#1: performance #1: (#2, #3), #4: performance #1: (#5, #6)}; or, the first information can be {#1: performance #1: (#2, #3), #4: performance #2: (#5, #6)}; or, the first information can be {#1: performance #1: (#2, #3), #4: {performance #1, performance #2}: (#5, #6)}; or, the first information can be {#1: {performance #1, performance #3}: (#2, #3), #4: {performance #1, performance #2}: (#5, #6)}.

[0208] Optionally, the first information also includes at least one performance metric associated with one or more performance metrics. The one or more performance metrics are some or all of the at least one performance metric within the first information. A performance metric can be associated with one or more performance metrics. In this case, different sets of IDs may correspond to the same or different performance metrics.

[0209] The IDs that use at least one of the above AI functions include #1 and #4, and at least one set of IDs can be an example of {(#2, #3), (#5, #6)}.

[0210] The first information can be {#1: performance #1: {performance index #1}: (#2, #3), #4: performance #1: {performance index #1}: (#5, #6)}; or, the first information can be {#1: performance #1: {performance index #1}: (#2, #3), #4: performance #1: {performance index #2}: (#5, #6)}; or, the first information can be {#1: performance #1: {performance index #1, performance index #2}: (#2, #3), #4: performance #1: {performance index #2}: (#5, #6)}; or, the first information can be {#1: performance #1: {performance index #1}(#2, #3), #4: performance #2: {performance index #1}(#5, #6)}; or, the first information can be {#1: performance #1: {performance index #1}(#2, #3), #4: performance #2: {performance index #1}(#5, #6)}; or, the first information can be {#1: performance #1: {performance index #1}(#2, #3), #4: performance #2: {performance index #1}(#5, #6)}; or, the first information can be {#1: performance #1: {performance index #1}(#2, #3), #4: performance #2: {performance index #1}(#5, #6)}; or, the first information can be {#1: performance #1: {performance index #1}(#2, #3), #4: performance #2: {performance index #1}(#5, #6)}; #1}(#2, #3), #4: Performance #2: {Performance Indicator #2, Performance Indicator #3}(#5, #6)}; or, the first information can be {#1: Performance #1: {Performance Indicator #1, Performance Indicator #3}(#2, #3), #4: Performance #2: {Performance Indicator #1}(#5, #6)}; the first information can be {#1: Performance #1: {Performance Indicator #1, Performance Indicator #2}(#2, #3), #4: {Performance #1, Performance #2}: {Performance Indicator #1, Performance Indicator #2}(#5, #6)}; or, the first information can be {#1: {Performance #1, Performance #3}: {Performance Indicator #1, Performance Indicator #2}(#2, #3), #4: {Performance #1, Performance #2}: {Performance Indicator #1, Performance Indicator #2}(#5, #6)}.

[0211] (2-3) The first information includes the ID of at least one AI function and at least one performance information associated with some or all of the IDs of the at least one AI function.

[0212] Different performance metrics of the same AI function may be related. For example, performance metrics #1 and #2 of AI function #1 are related, and performance metrics #1, #2, and #3 of AI function #2 are related. Taking this as an example, the first piece of information could be {#1: (performance metrics #1, performance metrics #2), #2: (performance metrics #1, performance metrics #2, performance metrics #3)}.

[0213] Different AI models under the same performance metric may have correlations. For example, for performance metric #1, the performance of AI function #1 and the performance of AI function #2 are correlated; for performance metric #2, the performance of AI function #1, the performance of AI function #2, and the performance of AI function #3 are correlated. Taking this as an example, the first piece of information could be {performance metric #1: (#1, #2), performance metric #2: (#1, #2, #3)}.

[0214] Optionally, the terminal device may report the applicability / applicability conditions of different AI functions to the network device separately; or, the terminal device may report the applicability / applicability conditions of multiple AI functions together to the network device. The applicability / applicability conditions and first information of at least one AI function are carried on different signaling messages. In other words, the applicability / applicability conditions and first information of at least one AI function are reported to the network device independently. Alternatively, the applicability / applicability conditions and first information of at least one AI function may be carried on a single signaling message. In other words, the applicability / applicability conditions and first information of at least one AI function may be reported to the network device together. For example, the first information may also include one or more of the following: applicability conditions associated with the AI ​​function, performance-related applicability conditions, or performance indicator-related applicability conditions.

[0215] For example, in (2-1), the first information may include at least one set of IDs, as well as the applicable conditions corresponding to at least one set of IDs. Taking the correlation between the performance of AI function #1 under applicable condition #1, the performance of AI function #2 under applicable condition #2, and the performance of AI function #3 under applicable condition #3 as an example, the first information may include {(applicable condition #1: #1), (applicable condition #2: #2), (applicable condition #3: #3)}.

[0216] In (2-1), the first information, in addition to including at least one set of IDs and performance information associated with one or more sets of IDs, may also include one or more applicable conditions associated with the performance. Taking the correlation of the performance #1 of AI function #1 under applicable condition #1, the performance #2 of AI function #2 under applicable condition #2, and the performance #1 of AI function #3 under applicable condition #3 as an example, the first information may include {(performance #1: applicable condition #1: #1), (performance #2: applicable condition #2: #2), (performance #1: applicable condition #3: #3)}. Taking the correlation of the performance index #1 of AI function #1 under applicable condition #1, the performance index #2 of AI function #2 under applicable condition #2, and the performance index #2 of AI function #3 under applicable condition #3 as an example, the first information may include {(performance #1: applicable condition #1: performance index #1: #1), (performance #1: applicable condition #2: performance index #2: #2), (performance #1: applicable condition #3: performance index #2: #3)}.

[0217] Similarly, in (2-2), the first information may include, in addition to the ID of at least one AI function and at least one set of IDs, one or more applicable conditions related to performance. For example, the performance #1 of AI function #1 under applicable condition #1 is associated with the performance #1 of AI function #2 under applicable condition #1 and the performance #1 of AI function #3 under applicable condition #2, respectively. However, the performance #1 of AI function #2 and the performance #1 of AI function #3 are not associated. The first information could be {applicable condition #1: performance #1: #1: (applicable condition #1: performance #1: #2, applicable condition #2: performance #1: #3)}.

[0218] In (2-3), the first information may include, in addition to the ID of at least one AI function and at least one performance information associated with the ID of the at least one AI function, one or more applicable conditions associated with the performance. Taking the relationship between the performance index #1 of AI function #1 under applicable condition #1 and the performance index #2 under applicable condition #2 as an example, the first information may be {#1: (applicable condition #1: performance index #1, applicable condition #2: performance index #2)}.

[0219] There are no restrictions on the specific implementation form of the information used to characterize the applicability conditions of AI functions. For example, the applicability conditions of AI functions include the type of terminal device. Accordingly, the information used to characterize the applicability conditions of AI functions may include the ID of the terminal device type. Taking the terminal device types as including V2X terminal devices, Unmanned Aerial Vehicle (UVA) terminal devices, and enhanced mobile broadband (eMBB) terminal devices, with the identifier of V2X terminal devices as ID1, UAV terminal devices as ID2, and eMBB terminal devices as ID3, if the terminal device type information includes ID1, it means that the terminal device capable of performing AI functions is a V2X terminal; if the terminal device type information includes both ID1 and ID2, it means that the terminal devices capable of performing AI functions are both V2X terminal devices and UAV terminal devices; if the terminal device type information includes ID3, it means that the terminal device capable of performing AI functions is an eMBB terminal device. As another example, the applicability conditions of AI functions include the computing power information of the terminal device. This information can be a computing power level or a specific computing power value.

[0220] S602, the performance of network devices in acquiring the first AI function.

[0221] The performance of the first AI function can be determined by the terminal device, which then sends the performance information of the first AI function to the network device. For example, the terminal device processes the input raw information using the AI ​​model corresponding to the first AI function to obtain information A; the terminal device determines the performance of the first AI function based on information A and sends the performance information to the network device. Taking the application of the first AI function to positioning as an example, the terminal device processes the input raw information (such as the target's starting position, moving speed, and moving direction) using the AI ​​model corresponding to the first AI function to obtain information A (such as the target's location information); the terminal device determines the performance of the first AI function based on the target's location information and the target's actual location; the terminal device then sends the performance information of the first AI function to the network device. For example, if the performance of the first AI function meets a first indicator, the terminal device sends information characterizing the performance of the first AI function to the network device.

[0222] Alternatively, the performance of the first AI function can be determined by the network device itself. For example, the terminal device processes the input raw information using the AI ​​model corresponding to the first AI function to obtain information A; the terminal device sends information A to the network device, and the network device determines the performance of the first AI function based on information A. Taking the application of the first AI function to positioning as an example, the terminal device processes the input raw information (such as the target's starting position, moving speed, moving direction, etc.) using the AI ​​model corresponding to the first AI function to obtain the target's position information; the terminal device sends the target's position information to the network device, and the network device determines the performance of the first AI function based on the target's position information obtained by the first AI function and the target's actual position.

[0223] S603. The network device determines the performance of the second AI function based on the performance of the first AI function and the first information.

[0224] The performance of the second AI function is related to the performance of the first AI function. The network device determines the performance of the first AI function and can determine the performance of the second AI function based on the performance of the first AI function and the first information. In this embodiment, the performance of an inactive AI function is mainly considered in terms of its applicability; for an activated AI function, the performance metrics are mainly considered. The following example, using the detection of the performance of the second AI function, illustrates the determination of its performance. In the following example, #N represents the AI ​​function ID N, for example, #1 is the ID of the first AI function, and #2 is the ID of the second AI function (this will be used as an example below).

[0225] In Example 3-1, it is assumed that the second AI function is an inactive AI function, and the first information includes the relationship between the applicable conditions of the first AI function and the applicable conditions of the second AI function.

[0226] For example, the first information includes {(Applicable Condition #1: #1):(Applicable Condition #1: #2)}. If the network device determines that the first AI function is applicable under Applicable Condition #1, then the network device can determine, based on the performance of the first AI function and the first information, that the second AI function is also applicable under Applicable Condition #1. The network device's determination that the first AI function is applicable under Applicable Condition #1 can be based on the applicability of the first AI function reported by the terminal device.

[0227] For example, the first information includes {(applicable condition #1: performance #1: #1):(applicable condition #1: performance #2: #2)}. If the network device determines that the first AI function is applicable under applicable condition #1 and the performance #1 is good, then the network device can determine that the second AI function is also applicable under applicable condition #1 and the performance #2 is also good, based on the performance of the first AI function and the first information.

[0228] For example, the first information includes {(applicable condition #1: performance index #1: #1):(applicable condition #1: performance index #2: #2)}. If the network device determines that the first AI function is applicable under applicable condition #1 and can achieve performance index #1, then the network device can determine, based on the performance of the first AI function and the first information, that the second AI function is also applicable under applicable condition #1 and can achieve performance index #2.

[0229] In Example 3-2, it is assumed that the second AI function is to activate the AI ​​function, and the first information includes the correlation between the performance (or performance index) of the first AI function and the performance (or performance index) of the second AI function.

[0230] For example, the first information includes {#1: performance #1: (#2, #3), #4: performance #1: (#5, #6)}. If the network device determines that the performance #1 of the first AI function is better, then the network device can determine that the performance #1 of the second AI function is also better based on the performance of the first AI function and the first information.

[0231] For example, the first information includes {#1: performance #1: {performance index #1}: (#2, #3), #4: performance #1: {performance index #1}: (#5, #6)}. If the network device determines that the performance #1 of the first AI function can achieve the performance index #1, then the network device can determine, based on the performance of the first AI function and the first information, that the performance #1 of the second AI function can also achieve the performance index #1.

[0232] A network may contain multiple AI functions. If the performance of each AI function is assessed individually, the resource and signaling overhead can be substantial. In this embodiment, for multiple AI functions with related performance, only the performance of one AI function needs to be tested to determine the performance of the other AI functions. This reduces the number of AI functions tested, thereby lowering resource and signaling overhead.

[0233] S604. The network device determines the management strategy for the second AI function based on its performance.

[0234] After determining the second AI function, the network device can determine the management policy for the second AI function, and then perform corresponding operations according to the determined management policy to manage the second AI function.

[0235] As mentioned earlier, AI function management strategies include activating, deactivating, updating, retraining, or switching AI functions. It should be understood that activating an AI function applies to inactive AI functions, deactivating it applies to activated AI functions, updating a portion of the AI ​​function's functionality, retraining it involves modifying the corresponding model parameters, and switching it means replacing one AI function with another. Activating an AI function can also be replaced by enabling the configuration to take effect. Deactivating an AI function can also be replaced by disabling the configuration to enable it.

[0236] Depending on the performance of the second AI function, the management strategy for the second AI function will also vary.

[0237] For example, if the second AI function is an inactive AI function, it will not be activated when it is no longer applicable; however, it can be activated when it is applicable. Alternatively, if the second AI function is inactive and no longer applicable, its management strategy could be to switch it, for example, switching it to a suitable AI function. For instance, when a network device activates the second AI function, it can execute S605: The network device can send a first message to the terminal device to activate the second AI function, carrying the ID of the second AI function. The specific name of the first message is not limited; for example, it could be called an AI function activation message. When the terminal device needs to switch from the second AI function to a third AI function, it can send a second message to the terminal device to switch the second AI function, carrying the ID of the third AI function. The specific name of the second message is not limited; for example, it could be called an AI function switching message.

[0238] For example, if the second AI function is an activated AI function, and its performance is poor, the management strategy for the second AI function could be to deactivate it, update it, or retrain it. Here, "good performance" means the second AI function meets preset performance indicators; "poor performance" means it does not meet these indicators. Optionally, when the network device deactivates the second AI function, it can send a first message to the terminal device. This first message carries the ID of the second AI function. The specific name of the first message is not limited; for example, it could be called an "AI function deactivation message."

[0239] In possible scenarios, terminal devices can report the required performance metrics for a second AI function to the network device to assist the network device in deciding on the management strategy for the AI ​​function. Taking the second AI function as an example, the terminal device can send second information to the network device, which can indicate that the first performance metric of the second AI function does not meet the first performance metric; the network device can then determine to deactivate the second AI function based on the second information. Alternatively, the second information can indicate that the first performance metric of the second AI function is the second performance metric, and if the second performance metric is lower than the first performance metric, the network device can determine that the first performance metric does not meet the first performance metric and therefore decide to deactivate the second AI function.

[0240] Optionally, the terminal device, triggered by the network device, reports the results of one or more performance metrics of the second AI function. For example, the network device may send third information to the terminal device, indicating whether the performance metrics of the second AI function have reached preset performance metrics. For example, the third information may include information about the preset performance metrics (e.g., 80%). If the performance metrics of the second AI function reach the preset performance metrics, the terminal device may send an acknowledgement (ACK) message to the network device; if the performance metrics of the second AI function reach the preset performance metrics, the terminal device may send a no-acknowledge (NACK) message to the network device, or the terminal device may choose not to send ACK / NACK messages within a preset time period. The network device determines the management strategy for the second AI function based on the feedback from the terminal device. For example, if the terminal device's feedback is ACK, then the management strategy for the second AI function may be to activate the second AI function.

[0241] Figure 6 illustrates an example where the network device is an access network device. In possible scenarios, the core network device can send a first request to the terminal device, the terminal device determines first information, sends the first information to the core network device, and the core network device then forwards the first information to the access network device. Alternatively, the core network device can manage at least one AI function of the terminal device, determine the first information independently, and send the first information to the access network device. Optionally, the core network device can instruct the access network device to configure parameters for the terminal device's AI function, or instruct the access network device to associate the terminal device's AI function.

[0242] In this embodiment, the terminal device can report AI functions with performance correlation to the network device. In this way, the network device can deduce the performance of one AI function based on the performance of another AI function, without having to test the performance of multiple AI functions separately. This reduces the number of AI functions to be tested, thereby reducing the resource and signaling overhead of AI function performance testing.

[0243] Example 2

[0244] The difference between Embodiment 2 and Embodiment 1 is that the network device determines the first information on its own and feeds the first information back to the terminal device, so that the terminal device can determine the performance of the second AI function based on the performance of the first AI function and the first information.

[0245] Please refer to Figure 7, which is a flowchart illustrating the first communication method provided in this embodiment. Figure 7 describes the method from the perspective of interaction between network devices and terminal devices. As shown in Figure 7, the communication method includes the following steps.

[0246] S700, network devices determine the first information.

[0247] For details regarding the first information, please refer to the description of the first information in the embodiment; it will not be repeated here.

[0248] In this embodiment, the network device can determine the first information itself. For example, the network device can obtain one or more historical information of multiple AI functions, such as: applicable environmental conditions, the capability of the device deploying the AI ​​function, the status of the device deploying the AI ​​function, or the configuration associated with the AI ​​function, and obtain the historical applicability, historical performance, historical performance indicators, etc. of these multiple AI functions, and determine the first information based on the obtained information.

[0249] S701, the network device sends first information to the terminal device, and correspondingly, the terminal device receives the first information from the network device.

[0250] Regarding the content of the first information, refer to the relevant description of S601 in the embodiment. The difference between S601 and S701 is that in S701, the first information is sent from the network device to the terminal device, and the first information can be carried in an RRC message, MAC CE, or DCI.

[0251] Optionally, the network device receives a first request from the terminal device and sends first information to the terminal device. The first request can be referred to in the description in S600 above, and will not be repeated here.

[0252] S702, The terminal device determines the performance of the second AI function based on the first AI function and the first information.

[0253] For details on S702, please refer to the aforementioned S602 where the network device determines the performance of the second AI function based on the first AI function and the first information; this will not be repeated here.

[0254] S703. The terminal device performs corresponding operations based on the performance of the second AI function.

[0255] Based on the performance of the second AI function, the terminal device can decide on the management strategy for the second AI function on its own, or it can request the network device to decide on the management strategy for the second AI function.

[0256] Depending on the performance of the second AI function, the corresponding operations will vary. For example, if the second AI function is not activated and the terminal device determines that it is not applicable, then the second AI function does not need to be activated, or it needs to be switched. If the second AI function is applicable, then it can be activated. If the second AI function is activated, but the terminal device determines that its performance is poor, then the second AI function does not need to be activated, or it needs to be updated, or it needs to be retrained.

[0257] Optionally, the terminal device executes S704: The terminal device informs the network device of the management policy for the second AI function. For example, the terminal device sends third information to the network device, which may indicate the management policy for the second AI function. The terminal device determines the management policy for the second AI function and may request the network device to execute the management policy. For example, if the management policy for the second AI function is to activate the second AI function, the terminal device may send a second request to the network device to request the network device to activate the second AI function. This second request may include the ID of the second AI function. As another example, if the management policy for the second AI function is to switch to a third AI function, the terminal device may send a second request to the network device to request the network device to switch the second AI function. This second request may include the ID of the third AI function. Based on the third information or the second request, the network device may determine whether to disable the configuration corresponding to the second AI function or send the configuration of the third AI function.

[0258] Optionally, the terminal device can inform the network device of the performance test results of the second AI function, assisting the network device in determining the management strategy for the second AI function. For example, the terminal device sends the performance indicators achieved by one or more of the second AI functions to the network device. If the performance indicators of one or more AI functions do not meet the preset performance indicators, then the network device determines to deactivate the second AI function.

[0259] In this embodiment, the network device can notify the terminal device of AI functions with performance correlation. In this way, the terminal device can deduce the performance of one AI function based on the performance of another AI function, thereby deciding on the management strategy of the corresponding AI function or requesting the network device to decide on the management strategy of the AI ​​function. For example, not activating AI functions with poor performance helps to save resource consumption in the network system.

[0260] The methods provided in the embodiments of this application above are described using network devices and terminal devices as examples. In this application, each embodiment can be implemented independently or in combination based on certain inherent connections; in each embodiment, different implementation methods can be implemented in combination or independently. To achieve the functions in the methods provided in the embodiments of this application above, access network devices and access network equipment may include hardware structures and / or software modules, implementing the above functions in the form of hardware structures, software modules, or a combination of hardware structures and software modules. Whether a particular function is implemented in the form of hardware structures, software modules, or a combination of hardware structures and software modules depends on the specific application and design constraints of the technical solution.

[0261] Based on the same inventive concept as the method embodiments, this application provides a communication device. The communication device used to implement the above method in the embodiments of this application is described below with reference to the accompanying drawings. The content above can be used in subsequent embodiments, and repeated content will not be repeated.

[0262] Figure 8 is a schematic block diagram of a communication device 800 provided in an embodiment of this application. This communication device 800 can correspondingly implement the functions or steps implemented by network devices or terminal devices in the various method embodiments described above. The communication device 800 may include modules or units that implement the methods described above; for example, the communication device 800 includes a processing unit 810 and a communication unit 820. Optionally, the communication device may also include a storage unit 830, which can be used to store instructions (code or program) and / or data. The storage unit may be, for example, a memory. The processing unit 810 and the communication unit 820 may be coupled to the storage unit. For example, the processing unit 810 can read instructions (code or program) and / or data from the storage unit to implement the corresponding method.

[0263] The processing unit 810 may be a processor or controller, such as a CPU, general-purpose processor, DSP, ASIC, FPGA, programmable logic device (PLD), or other programmable logic device, transistor logic device, hardware component, or any combination thereof. It can implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. The processor may also be a combination that implements computing functions, such as including one or more microprocessor combinations, a combination of DSP and microprocessor, etc. When the communication device 800 is a system-on-a-chip, the processing unit 810 may be an FPGA, a dedicated ASIC, a SoC, a CPU, a network processor (NP), a DSP, an MCU, a PLD, or other integrated chip. The processing unit 810 may be a processor containing a system-on-a-chip.

[0264] The communication unit 820 is a transceiver, interface circuit, bus, pin, or other possible communication interface used to receive signals from other devices. For example, when the device is implemented as a chip, the communication unit 820 is an interface circuit for the chip to receive signals from other chips or devices, or an interface circuit for the chip to send signals to other chips or devices. For example, the interface circuit can be a code / data read / write interface circuit. This interface circuit can be used to receive code instructions (stored in memory, which can be read directly from memory or through other devices) and transmit them to the processing unit 810. As another example, the interface circuit can also be a signal transmission interface circuit between a communication processor and a transceiver. When the communication device 800 is a chip-type device or circuit, the communication unit 820 can be an input / output circuit and / or a communication interface; the processing unit 810 is an integrated processor, microprocessor, or integrated circuit.

[0265] When the communication device 800 is a module or unit within a device, the storage unit 830 can be a storage unit within the chip, such as a register or cache. For example, the storage unit 830 can also be a storage unit located outside the chip within a terminal device or access network device, such as read-only memory (ROM) or other types of static storage devices capable of storing static information and instructions, such as random access memory (RAM), flash memory, or hard disk. The aforementioned units can be configured independently or partially or completely integrated. In one example, computer program instructions for executing the above embodiments can be stored in non-volatile memory, such as at least a portion of the aforementioned memory 830 (e.g., one or more of ROM, flash memory, or hard disk). When the terminal device or access network device is running, the corresponding computer program instructions may be partially or wholly loaded into a memory with a faster transmission speed than the processing unit 810, such as at least a portion of the memory 830 described above (e.g., one or more of RAM, cache, or register), for the processing unit 810 to execute in order to implement the steps in the above method embodiments.

[0266] In one implementation, the communication device 800 can correspondingly implement the behavior and functions of the terminal device in the above method embodiments. The communication device 800 can be the terminal device, a component within the terminal device (e.g., a module, chip, or circuit), a part of a module, chip, or chipset in the terminal device used to execute the relevant method functions, or a software module in the terminal device capable of implementing the above communication method; there are no limitations. For details, please refer to the relevant content of the foregoing method embodiments, which will not be repeated here.

[0267] For example, the communication unit 820 is used to receive first information, which indicates a correlation between the performance of the first AI function and the performance of the second AI function. The processing unit 810 is used to obtain the performance of the first AI function and determine the performance of the second AI function based on the performance of the first AI function and the first information.

[0268] As an optional implementation, the performance of two AI functions is related if they satisfy one or more of the following: the performance of one AI function is better than or equal to the performance of the other AI function; or, the applicable conditions of the two AI functions are related, and the applicable conditions include one or more of the following: applicable environmental conditions, the capability of the device deploying the AI ​​function, the status of the device deploying the AI ​​function, and the associated configuration of the AI ​​functions.

[0269] As an optional implementation, the two AI functions are a first AI function and a second AI function. The applicable conditions of the two AI functions are related, including: the applicable conditions of the first performance of the first AI function and the applicable conditions of the second performance of the second AI function are related; or, the applicable conditions of the first performance index of the first performance of the first AI function and the applicable conditions of the second performance index of the second performance of the second AI function are related.

[0270] As an optional implementation, the first information includes: at least one set of IDs, wherein one set includes the IDs of performance-related AI functions.

[0271] As an optional implementation, at least one ID set includes a first ID set, and the first information also includes performance information associated with some or all of the IDs in the first ID set.

[0272] As an optional implementation, the first information includes at least one AI function ID and at least one set of IDs, wherein one set of IDs corresponds to one ID, including the IDs of AI functions associated with the performance of the AI ​​function indicated by the corresponding ID.

[0273] As an optional implementation, at least one AI function ID includes a first ID, at least one ID set includes a first ID set, the first ID and the first ID set correspond, and the first information also includes performance information associated with the first ID and / or performance information associated with some or all of the IDs in the first ID set.

[0274] As an optional implementation, the first information includes: the ID of at least one AI function and at least one performance information associated with some or all of the IDs of the at least one AI function.

[0275] As an optional implementation, the first information also includes at least one performance metric associated with each performance metric.

[0276] As an optional implementation, the processing unit 810 is also configured to: determine a management strategy for the second AI function based on the performance of the second AI function, the management strategy including one or more of the following: activating the second AI function, deactivating the second AI function, updating the second AI function, retraining the second AI function, or switching the second AI function.

[0277] As an optional implementation, the communication unit 820 is also used to send a first request for an AI function whose performance is correlated.

[0278] In one implementation, the communication device 800 can correspondingly implement the behavior and functions of the network device in the above method embodiments. The communication device 800 can be a network device, a component within the network device (e.g., a module, chip, or circuit), a part of a chip or chipset within the network device used to execute the relevant method functions, or a software module within the network device capable of implementing the above communication method; there are no limitations. For details, please refer to the relevant content of the foregoing method embodiments, which will not be repeated here.

[0279] For example, processing unit 810 is used to determine first information, which indicates a correlation between the performance of a first AI function and the performance of a second AI function. Communication unit 820 is used to send the first information.

[0280] As an optional implementation, the performance of two AI functions is related if they satisfy one or more of the following: the performance of one AI function is better than or equal to the performance of the other AI function; or, the applicable conditions of the two AI functions are related, which include one or more of the following: applicable environmental conditions, the capability of the device deploying the AI ​​function, or the configuration associated with the AI ​​function.

[0281] As an optional implementation, the two AI functions are a first AI function and a second AI function. The applicable conditions of the two AI functions are related, including: the applicable conditions of the first performance of the first AI function and the applicable conditions of the second performance of the second AI function are related; or, the applicable conditions of the first performance index of the first performance of the first AI function and the applicable conditions of the second performance index of the second performance of the second AI function are related.

[0282] As an optional implementation, the first information includes: at least one set of IDs, wherein one set includes the IDs of performance-related AI functions.

[0283] As an optional implementation, at least one ID set includes a first ID set, and the first information also includes performance information associated with some or all of the IDs in the first ID set.

[0284] As an optional implementation, the first information includes at least one AI function ID and at least one set of IDs, wherein one set of IDs corresponds to one ID, including the IDs of AI functions associated with the performance of the AI ​​function indicated by the corresponding ID.

[0285] As an optional implementation, at least one AI function ID includes a first ID, at least one ID set includes a first ID set, the first ID and the first ID set correspond, and the first information also includes performance information associated with the first ID and / or performance information associated with some or all of the IDs in the first ID set.

[0286] As an optional implementation, the first information includes: the ID of at least one AI function and at least one performance information associated with some or all of the IDs of the at least one AI function.

[0287] As an optional implementation, the first information also includes at least one performance metric associated with each performance metric.

[0288] As an optional implementation, the communication unit 820 is also used to: receive a first request for an AI function whose performance is related.

[0289] When the communication device 800 is a chip-based device or circuit, the transceiver module can be an input / output circuit and / or a communication interface; the processing module is an integrated processor, microprocessor, or integrated circuit.

[0290] In one possible design, when the communication device 800 is a network device or a terminal device; or when the communication device 800 is a communication module within a network device or a terminal device, the function of the processing unit 810 can be implemented by one or more processors. Specifically, the processor may include a modem chip, or a system-on-a-chip (SoC) chip or a SIP chip containing a modem core. The function of the communication unit 820 can be implemented by a transceiver circuit.

[0291] In one possible design, when the communication device 800 is a circuit or chip responsible for communication functions in a network device or terminal device, such as a modem chip or a system-on-a-chip (SoC) or SIP chip containing a modem core, the function of the processing unit 810 can be implemented by a circuit system in the aforementioned chip that includes one or more processors or processor cores. The function of the communication unit 820 can be implemented by the interface circuitry or data transceiver circuitry on the aforementioned chip.

[0292] In one possible design, when the communication device 800 is a network device or a terminal device; or when the communication device 800 is a communication and / or computing module within a network device or terminal device, the function of the processing unit 810 can be implemented by one or more processors. Specifically, the processor may include a GPU, or a system-on-a-chip (SoC) or SIP chip containing a GPU. Alternatively, the processor may include an AI processor, or a SoC or SIP chip containing an AI processor. Alternatively, the processor may include an ASIC, or a SoC or SIP chip containing an ASIC. Alternatively, the processor may include an NPU, or a SoC or SIP chip containing an NPU. The function of the communication unit 820 can be implemented by transceiver circuitry.

[0293] In one possible design, when the communication device 800 is a circuit or chip responsible for communication and / or computing functions in a network device or terminal device, such as a GPU or a system-on-a-chip (SoC) or SIP chip containing a GPU, an AI processor or a SoC or SIP chip containing an AI processor, an NPU or a SoC or SIP chip containing an NPU, or an ASIC or a SoC or SIP chip containing an ASIC, the function of the processing unit 810 can be implemented by a circuit system in the aforementioned chip that includes one or more processors or processor cores. The function of the communication unit 820 can be implemented by interface circuits or data transceiver circuits on the aforementioned chip.

[0294] It is understood that the division of units in the aforementioned communication device 800 is merely a logical functional division. One function can correspond to one functional unit, or two or more functions can be integrated into one functional unit. In actual implementation, all or some units can be integrated into one physical entity, or distributed across different physical entities. Furthermore, the aforementioned functional units can be implemented in hardware, software, or a combination of both. Whether a function is executed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for specific applications, but such implementation should not be considered beyond the scope of this application.

[0295] In one example, the functional unit in any of the above devices may be one or more integrated circuits configured to implement the above methods, such as: one or more ASICs, or one or more CPUs, one or more microcontroller units (MCUs), one or more DSPs, or one or more FPGAs, or a combination of at least two of these integrated circuit forms.

[0296] Figure 9 is a schematic block diagram of a terminal device 900 provided in an embodiment of this application. This terminal device 900 corresponds to the terminal device in Figure 5 and is used to implement the operation of the terminal device in the above embodiments. Specific functions can be found in the descriptions of the above method embodiments. As shown in Figure 9, the terminal device 900 includes: one or more antennas 910, a radio frequency processing system 920, and a processor system 930.

[0297] In the downlink or sidelink direction, the RF processing system 920 receives RF signals through the antenna 910 and sends the RF-processed signals to the processor system 930 for further processing. In the uplink or sidelink direction, the processor system 930 processes the terminal-side information and sends it to the RF processing system 920, which then processes the signal and transmits it through the antenna 910.

[0298] In one example, the radio frequency (RF) processing system 920 serves as the communication interface for external communication of the terminal device and may include an RF front end (RFFE) 921 and an RF transceiver 922. The RFFE 921 is primarily used for one or more processing operations, such as shaping, passband selection, or gain adjustment, on the RF signals received by the antenna or those to be transmitted through the antenna. It may include one or more components such as RF switches, duplexers, filters, power amplifiers, antenna tuners, and low-noise amplifiers. The RFFE 921 can be a circuit system composed of multiple discrete components or integrated into one or more chips. The RF transceiver 922 processes the RF signals received by the RFFE into baseband / IF signals for further processing by the processor system 930, and processes the baseband / IF signals provided by the processor system 930 into RF signals for transmission to the RFFE 921. The baseband / IF signals transmitted between the RF transceiver 922 and the processor system 930 can be digital or analog signals. The RF transceiver 922 can be implemented by one or more chips, which are commonly referred to as RF ICs.

[0299] In one example, the processor system 930 may include one or more processors for processing signals and executing one or more communication protocols. Optionally, the processor system 930 may also include a memory 936. In one example, the one or more processors include at least one baseband processor 931 (also known as a modem processor). The memory 936 is used to store data and / or computer program instructions. Optionally, the processor system 930 may also include one or more application processors 932 for implementing processing of the terminal operating system and application layer. The application processor 932 may include, for example, a GPU, NPU, AI processor, or ASIC. Optionally, the processor system 930 may also include one or more of a voice subsystem 933, a multimedia subsystem 934, or an interface circuit 935. The voice subsystem 933 is used to process voice signals, the multimedia subsystem 934 is used to handle multimedia-related operations, such as video encoding / decoding, image processing, etc., and the interface circuit 935 is used to enable communication with other terminal components, such as a display 940, an input device 950, a memory 960, etc. The above-mentioned components in the processor system 930 can communicate with each other via a bus or communication interface circuit.

[0300] In one example, the processor system 930 can be packaged as a single processor chip, such as a SoC chip or a SIP chip. In another example, the processor system 930 can be a system composed of multiple chips; for example, the baseband processor 931 can be packaged as a single chip, or packaged with part or all of the circuitry of the radio frequency processing system into a single chip.

[0301] In one example, memory 936 can be on-chip memory, i.e., located on the system-on-a-chip (SoC) 930. In another example, memory 960 can be off-chip memory, i.e., located outside the SoC 930.

[0302] In one example, the baseband processor 931 may include one or more processor cores 9311 and interface circuitry 9314. The one or more processor cores 9311 are used to process signals and execute one or more communication protocols. Optionally, the baseband processor 931 may also include a memory 9312 for storing at least a portion of the corresponding computer program instructions and / or data. In one example, the one or more processor cores 9311 execute the computer program instructions stored in the memory 9312 to perform the relevant operations (such as generating a DCI, generating at least one TB) described in the above method embodiments. In this disclosure, the memory 9312 is used to store corresponding computer program instructions and / or data. This can mean that the memory 9312 stores all corresponding computer program instructions and / or data for execution by the processor core 9311; or it can mean that the memory 9312 stores a portion of the corresponding computer program instructions and / or data, including the computer program instructions and / or data currently required to be executed by the processor core 9311. The memory 9312 can store different portions of computer program instructions and / or data multiple times for execution by the processor core 9311 to implement the relevant operations in the above method embodiments. The interface circuit 9314 serves as a communication interface for communication with other components, such as transmitting signals with the radio frequency processing system 920, communicating with other subsystems and related components of the processor system 930 via a bus, such as transmitting data control signals with the application processor 932, and transmitting data or computer program instructions with the memory 936 or memory 960. Optionally, in order to reduce the load on the processor core, a baseband signal processing circuit 9313 can be set to perform at least some baseband signal processing, including one or more of signal demodulation, modulation, encoding or decoding.

[0303] In one example, the communication device provided in this application may be a terminal device 900, including a communication module comprising a processor system 930 and a radio frequency system 920, or a baseband processor 931.

[0304] The processor, processor system, application processor, baseband processor, processor circuit or processor core mentioned above can be collectively referred to as a processor. The processor may include one or more of the following: CPU, DSP, microprocessor unit (MPU), MCU, GPU, FPGA, ASIC, AI processor or NPU.

[0305] The aforementioned memory may include one or more of the following storage media: such as RAM, static random access memory (SRAM), dynamic random access memory (DRAM), phase-change memory (PCM), resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), cache, register, ROM, flash memory, erasable programmable read-only memory (EPROM), hard disk, etc. In one example, computer program instructions for executing the above embodiments may be stored on non-volatile memory, such as at least a portion of the aforementioned memory 960 (e.g., one or more of ROM, flash memory, EPROM, or hard disk). When the terminal is running, the corresponding computer program instructions may be partially or wholly loaded onto a memory with a faster transfer speed than the processor, such as at least a portion of memory 936 and / or memory 9312 (e.g., one or more of RAM, SRAM, DRAM, PCM, RERAM, MRAM, FRAM, cache, or register), for the processor to execute in order to implement the steps in the above method embodiments.

[0306] In one example, the RF transceiver 922 and the RF front-end 921 can also be packaged in a single chip. In another example, the RF transceiver 922, the RF front-end 921, and the baseband processor 931 can also be packaged in a single chip.

[0307] This application also provides a communication system, which includes at least one terminal device and a network device. The terminal device is a device for implementing the terminal device-related functions in the above-described communication method, and the network device is an access network device for implementing the network device-related functions in the above-described communication method.

[0308] This application also provides a computer-readable storage medium including instructions that, when run on a computer, cause the method executed by the terminal device or network device in the above-described communication method to be executed.

[0309] This application also provides a computer program product, including computer program code, which, when executed, causes the method executed by the terminal device or network device in the above-described communication method to be executed.

[0310] This application provides a chip system that includes a processor and may also include a memory, for implementing the functions of a network device or a terminal device in the aforementioned communication method.

[0311] To achieve the functions of the communication devices shown in Figures 8 and 9, this application embodiment also provides a chip, including a processor, for supporting the communication device in implementing the functions involved in the terminal device or network device in the above method embodiments. In one possible design, the chip is connected to a memory or the chip includes a memory for storing necessary computer programs, instructions, and data for the communication device.

[0312] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0313] Those skilled in the art will recognize that the various illustrative logical blocks and steps described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0314] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0315] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0316] The units described as separate components may or may not be physically separate. 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 the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0317] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the essential contributing part of the technical solution of this application, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0318] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A communication method, characterized in that, include: Receive first information, which indicates that the performance of the first artificial intelligence (AI) function is related to the performance of the second AI function. Obtain the performance of the first AI function; The performance of the second AI function is determined based on the performance of the first AI function and the first information.

2. The method as described in claim 1, characterized in that, If two AI functions satisfy one or more of the following conditions, then the performance of the two AI functions is correlated: The performance of one of the two AI functions is better than or equal to the performance of the other AI function; or... The applicable conditions of the two AI functions are related, and the applicable conditions include one or more of the following: applicable environmental conditions, the capability of the device deploying the AI ​​function, the status of the device deploying the AI ​​function, or the configuration associated with the AI ​​function.

3. The method as described in claim 2, characterized in that, The two AI functions are the first AI function and the second AI function, and the applicable conditions of the two AI functions are related, including: The applicable conditions of the first performance of the first AI function and the applicable conditions of the second performance of the second AI function are related; or, The applicable conditions of the first performance index of the first performance of the first AI function and the applicable conditions of the second performance index of the second performance of the second AI function are related.

4. The method according to any one of claims 1-3, characterized in that, The first information includes: At least one set of identifier IDs, wherein one set includes IDs of AI functions associated with performance.

5. The method as described in claim 4, characterized in that, The at least one ID set includes a first ID set, and the first information further includes performance information associated with some or all of the IDs in the first ID set.

6. The method according to any one of claims 1-3, characterized in that, The first information includes: At least one AI function ID and at least one set of IDs, wherein a set of IDs corresponds to one ID, including the IDs of AI functions associated with the performance of the AI ​​function indicated by the corresponding ID.

7. The method as described in claim 6, characterized in that, The ID of the at least one AI function includes a first ID, the at least one ID set includes a first ID set, the first ID and the first ID set correspond to each other, and the first information further includes: performance information associated with the first ID, and / or, performance information associated with some or all of the IDs in the first ID set.

8. The method according to any one of claims 1-3, characterized in that, The first information includes: The ID of at least one AI function, and at least one performance information associated with some or all of the IDs of the at least one AI function.

9. The method as described in claim 5, 7, or 8, characterized in that, The first information also includes: at least one performance metric associated with each performance.

10. The method according to any one of claims 1-9, characterized in that, The method further includes: Based on the performance of the second AI function, a management strategy for the second AI function is determined. The management strategy includes one or more of the following: activating the second AI function, deactivating the second AI function, updating the second AI function, retraining the second AI function, or switching the second AI function.

11. The method according to any one of claims 1-10, characterized in that, The method further includes: Send a first request, which is used to request AI functions that are related to performance.

12. A communication method, characterized in that, include: Determine first information, which is used to indicate a correlation between the performance of the first artificial intelligence (AI) function and the performance of the second AI function; Send the first message.

13. The method as described in claim 12, characterized in that, If two AI functions satisfy one or more of the following conditions, then the performance of the two AI functions is correlated: The performance of one of the two AI functions is better than or equal to the performance of the other AI function; or... The applicable conditions of the two AI functions are related, and the applicable conditions include one or more of the following: applicable environmental conditions, the capability of the device deploying the AI ​​function, the status of the device deploying the AI ​​function, or the configuration associated with the AI ​​function.

14. The method as described in claim 13, characterized in that, The two AI functions are the first AI function and the second AI function, and the applicable conditions of the two AI functions are related, including: The applicable conditions of the first performance of the first AI function and the applicable conditions of the second performance of the second AI function are related; or, The applicable conditions of the first performance index of the first performance of the first AI function and the applicable conditions of the second performance index of the second performance of the second AI function are related.

15. The method according to any one of claims 12-14, characterized in that, The first information includes: At least one set of identifier IDs, wherein one set includes IDs of AI functions associated with performance.

16. The method as described in claim 15, characterized in that, The first information also includes: performance information associated with some or all of the IDs in the first ID set.

17. The method according to any one of claims 12-14, characterized in that, The first information includes: At least one AI function ID and at least one set of IDs, wherein a set of IDs corresponds to one ID, including the IDs of AI functions associated with the performance of the AI ​​function indicated by the corresponding ID.

18. The method as described in claim 17, characterized in that, The ID of the at least one AI function includes a first ID, the at least one ID set includes a first ID set, the first ID and the first ID set correspond to each other, and the first information further includes: performance information associated with the first ID, and / or, performance information associated with some or all of the IDs in the first ID set.

19. The method according to any one of claims 12-14, characterized in that, The first information includes: The ID of at least one AI function, and the ID of at least one performance associated with some or all of the IDs of the at least one AI function.

20. The method as described in claim 16, 18, or 19, characterized in that, The first information also includes: at least one performance metric associated with each performance.

21. The method according to any one of claims 12-20, characterized in that, The method further includes: Receive a first request, which is used to request AI functions that are related to performance.

22. A communication device, characterized in that, The communication device includes a unit for performing the method as described in any one of claims 1 to 11, or includes a unit for performing the method as described in any one of claims 12 to 21.

23. A communication device, characterized in that, The communication device includes at least one processor, the at least one processor being configured to cause the method of any one of claims 1 to 11 to be executed by the communication device, or the at least one processor being configured to cause the communication device to execute the method of any one of claims 12 to 21.

24. A communication system, characterized in that, It includes a first device and a second device, the first device being used to perform the method as described in any one of claims 1 to 11, and the second device being used to perform the method as described in any one of claims 12 to 21.

25. A chip or chip system, characterized in that, The chip or chip system includes: At least one processor and an interface, the at least one processor being configured to call and execute instructions from the interface, such that when the at least one processor executes the instructions, the method as claimed in any one of claims 1 to 11 is executed, or the method as claimed in any one of claims 12 to 21 is executed.

26. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store a computer program that, when run on a computer, causes the method as described in any one of claims 1 to 11 to be performed, or causes the method as described in any one of claims 12 to 21 to be performed.

27. A computer program product, characterized in that, The computer program product includes a computer program that, when run on a computer, causes the method as described in any one of claims 1 to 11 to be performed, or causes the method as described in any one of claims 12 to 21 to be performed.