Communication method and related apparatus

Abstract

A communication method and a related apparatus, which are applied to the field of communications. The method comprises: acquiring learning-related state information associated with an artificial intelligence (AI) task, wherein the learning-related state information comprises one or more pieces of resource state information, environment state information, model performance information and data set state information; and sending the learning-related state information, wherein the learning-related state information is used for performing life cycle management on a model associated with the AI task. In the present solution, by means of performing operations, such as reporting learning-related state information, on a specific task, the impact of at least one of factors such as a resource state, an environment state, model performance, a data set state can be taken into consideration as much as possible when life cycle management is performed on a model associated with an AI task, thereby improving the efficiency and accuracy of the life cycle management for the model.

Description

A communication method and related device Technical Field

[0001] The present application relates to the field of communications, and in particular to a communication method and related devices. Background Art

[0002] To meet the vision of intelligent and inclusive future services, intelligence will further evolve within wireless network architectures. Artificial intelligence (AI) will be further integrated with wireless networks, enabling endogenous network intelligence. This will also include the development of intelligent terminals to address new demands and scenarios, including diversified terminal types, more flexible and intelligent terminal connections, and the incorporation of AI capabilities within terminals. Intrinsic network intelligence: In addition to traditional communication and connectivity services, networks will also provide computing and AI services to better support inclusive, real-time, and highly secure AI services. These new demands and scenarios will bring about changes in wireless network architecture and communication models.

[0003] Therefore, after the integration of AI and wireless networks, how to better improve the efficiency and accuracy of model lifecycle management in communication systems is a hot topic being studied by technical personnel in this field.

[0004] Summary of the Invention

[0005] The embodiments of the present application provide a communication method and related devices that can perform operations such as reporting learning-related status information for specific tasks, so that when performing lifecycle management of models associated with AI tasks, the influence of at least one of factors such as resource status, environmental status, model performance, and dataset status can be taken into account as much as possible, thereby improving the efficiency and accuracy of model lifecycle management.

[0006] In a first aspect, embodiments of the present application provide a communication method that can be performed by an apparatus. The apparatus can be a device (such as a terminal device or a network device), or a component of a device (such as a chip or a chip system or a circuit), which is not limited in this application.

[0007] The method comprises:

[0008] Obtaining learning-related status information associated with an artificial intelligence (AI) task, wherein the learning-related status information includes one or more of resource status information, environment status information, model performance information, and dataset status information;

[0009] The learning-related status information is sent, wherein the learning-related status information is used to perform lifecycle management on the model associated with the AI ​​task.

[0010] In the above method, by performing operations such as reporting learning-related status information for specific tasks, the lifecycle management of the model associated with the AI ​​task can take into account the influence of at least one of the factors such as resource status, environmental status, model performance, and dataset status as much as possible, thereby improving the efficiency and accuracy of model lifecycle management.

[0011] In a possible implementation manner, before acquiring the learning-related status information, the method further includes:

[0012] Receive indication information, wherein the indication information is used to indicate a sending method of the learning-related status information.

[0013] In another possible implementation, the indication information includes a reference time; and the sending of the learning-related status information includes:

[0014] When the reference time is met, the learning-related status information is sent.

[0015] In the above method, the learning-related status information may be sent in a timed manner by realizing a timer setting condition.

[0016] In another possible implementation, the instruction information includes a trigger condition; and the sending of the learning-related status information includes:

[0017] When the trigger condition is met, the learning-related status information is sent.

[0018] In the above method, the learning-related status information may be sent by achieving a set trigger condition.

[0019] In yet another possible implementation, the method further includes:

[0020] Sending indication information of a task identifier, wherein the indication information of the task identifier is used to indicate the associated AI task.

[0021] In the above method, the task identifier may be carried in the learning-related state information or may be independent. In another possible implementation, the resource state information includes resource margin, and the resource margin includes computing power margin and / or storage margin.

[0022] In the above method, the specific content of the resource status information is reflected.

[0023] In another possible implementation, the resource status information includes multiple task identifiers and multiple resource margins, and each of the multiple task identifiers is used to represent an AI task associated with one of the multiple resource margins.

[0024] In the above method, the content of resource status information associated with multiple AI tasks is reflected.

[0025] In another possible implementation, the resource status information is carried in the signaling of the Medium Access Control-Control Element (MAC-CE), and the signaling of the MAC-CE includes a task identification field and a computing power headroom (CPH) field, the task identification field includes the task identification, and the CPH field includes the resource margin.

[0026] In the above method, the field format of the MAC-CE signaling carrying the resource status information is reflected.

[0027] In yet another possible implementation, the method further includes:

[0028] Sending first indication information, wherein the first indication information is used to indicate that the learning-related state information comes from a first category of AI tasks or a second category of AI tasks, the first category of AI tasks includes one or more preset tasks, and the second category of AI tasks includes one or more AI tasks other than the first category of AI tasks.

[0029] In the above method, the first type of task can be one or more tasks preset by the standard or system, such as CSI compression feedback, beam prediction, positioning enhancement, and transceiver data transmission. Users of the communication system may subsequently use the second type of AI tasks. However, considering that the AI ​​tasks involved in each terminal, network device, or each area are different, a general learning-related state is set for the second type of tasks, namely the public learning-related state, to improve the reporting efficiency of learning-related state information and save resources.

[0030] In another possible implementation, the first indication information is a task identifier.

[0031] In the above method, the first indication information may be used to indicate whether the learning-related status information comes from a specific task alone, or may be used in addition to indicating whether the learning-related status information comes from a specific task or a public task.

[0032] In another possible implementation, the environmental status information includes first parameter information, where the first parameter information includes one or more of a system parameter, a scene type, and a model identifier.

[0033] In the above method, the specific content of the environmental status information is reflected.

[0034] In another possible implementation, the environmental status information includes multiple task identifiers and multiple first parameter information, and each task identifier in the multiple task identifiers is used to represent an AI task associated with one first parameter information in the multiple first parameter information.

[0035] In the above method, the content of the environmental state information associated with multiple AI tasks is reflected.

[0036] In another possible implementation, the environmental status information is carried in MAC-CE signaling, and the MAC-CE signaling includes a task identification field, a system parameter field, a scenario type field, and a model identification field, the task identification field includes the task identification, the system parameter field includes the system parameters, the scenario type field includes the scenario type, and the model identification field includes the model identification.

[0037] In the above method, the field format of the MAC-CE signaling carrying the environmental status information is reflected.

[0038] In another possible implementation, the model performance information includes second parameter information, which includes one or more of a model identifier, model performance, fallback performance, and fallback indication. The model performance is used to characterize the performance of the model used for the associated AI task, and the fallback performance is used to characterize the system performance of the associated AI task after the model execution falls back to non-model execution. The fallback indication is used to characterize whether the associated AI task needs to be executed back, and the model identifier includes the model identifier of one or more models in the model associated with executing the AI ​​task.

[0039] In the above method, the specific content of the model performance information is reflected.

[0040] In another possible implementation, the model performance information includes multiple task identifiers and multiple second parameter information, and each task identifier in the multiple task identifiers is used to represent an AI task associated with a second parameter information in the multiple second parameter information.

[0041] In the above method, the content of model performance information related to multiple AI tasks is reflected.

[0042] In another possible embodiment, the model performance information is carried in MAC-CE signaling, and the MAC-CE signaling includes a task identification field, a model identification field, a model performance field, a fallback performance field and a fallback indication field, the task identification field includes the task identification, the model identification field includes the model identification, the model performance field includes the model performance, the fallback performance field includes the fallback performance, and the fallback indication field includes the fallback indication.

[0043] In the above method, the field format of the MAC-CE signaling that carries the model performance information is reflected.

[0044] In another possible embodiment, the data set status information includes third parameter information, and the third parameter information includes one or more of data set size, data collection time, data collection location and data distribution description, wherein the data collection time includes the collection time of the latest data sample in the data set collected for the associated AI task and the collection time of the oldest data sample in the data set, the data set size is used to characterize the number of data samples in the data set collected for the associated AI task, the data collection location is used to characterize the location of the data in the data set collected for the associated AI task, and the data distribution description is used to characterize the distribution of data in the data set collected for the associated AI task.

[0045] In the above method, the specific content of the data set status information is reflected.

[0046] In another possible implementation, the data set status information includes multiple task identifiers and multiple third parameter information, and each task identifier in the multiple task identifiers is used to represent the AI ​​task associated with one third parameter information in the multiple third parameter information.

[0047] In the above method, the content of the data set status information associated with multiple AI tasks is reflected.

[0048] In another possible implementation, the data set status information is carried in MAC-CE signaling, and the MAC-CE signaling includes a task identification field, a data set size field, a latest data acquisition time field, an oldest data acquisition time field, a data acquisition location field, and a data distribution description field. The task identification field includes the task identification, the data set size field includes the data set size, the latest data acquisition time field includes the latest data acquisition time, the oldest data acquisition time field includes the oldest data acquisition time, the data acquisition location field includes the data acquisition location, and the data distribution description field includes the data distribution description.

[0049] In the above method, the field format of the MAC-CE signaling carrying the data set status information is reflected.

[0050] In a second aspect, the present application further provides a communication method, which corresponds to the communication method described in the first aspect and can be performed by an apparatus. The apparatus can be a device (such as a terminal device or a network device), or it can be a component of a device (such as a chip or a chip system or a circuit), which is not limited in this application. The method includes:

[0051] Receiving learning-related status information associated with the AI ​​task, wherein the learning-related status information includes one or more of resource status information, environment status information, model performance information, and dataset status information;

[0052] Perform lifecycle management on the model associated with the AI ​​task based on the learning-related status information.

[0053] In the above method, the efficiency and accuracy of lifecycle management of models associated with AI tasks can be improved based on the received learning-related status information of the associated AI tasks.

[0054] In a possible implementation manner, before receiving the learning-related status information, the method further includes:

[0055] Sending instruction information, wherein the instruction information is used to indicate a sending method of the learning-related status information.

[0056] In another possible implementation, the indication information includes a reference time, and the indication information is used to indicate that the learning-related status information should be sent when the reference time is met.

[0057] In the above method, the learning-related status information may be sent in a timed manner by realizing a timer setting condition.

[0058] In another possible implementation, the indication information includes a trigger condition, and the indication information is used to indicate that the learning-related status information should be sent when the trigger condition is met.

[0059] In the above method, the learning-related status information may be sent by achieving a set trigger condition.

[0060] In yet another possible implementation, the method further includes:

[0061] Receive indication information of a task identifier, wherein the indication information of the task identifier is used to indicate the associated AI task.

[0062] In the above method, the task identifier can be carried in the learning-related state information or can be independent.

[0063] In another possible implementation, the resource status information includes resource margin, and the resource margin includes computing power margin and / or storage margin.

[0064] In the above method, the specific content of the resource status information is reflected.

[0065] In another possible implementation, the resource status information includes multiple task identifiers and multiple resource margins, and each of the multiple task identifiers is used to represent an AI task associated with one of the multiple resource margins.

[0066] In the above method, the content of resource status information associated with multiple AI tasks is reflected.

[0067] In another possible implementation, the resource status information is carried in MAC-CE signaling, and the MAC-CE signaling includes a task identification field and a CPH field, the task identification field includes the task identification, and the CPH field includes the resource margin.

[0068] In the above method, the field format of the MAC-CE signaling carrying the resource status information is reflected.

[0069] In yet another possible implementation, the method further includes:

[0070] Receive first indication information, wherein the first indication information is used to indicate that the learning-related state information comes from a first category of AI tasks or a second category of AI tasks, the first category of AI tasks includes one or more preset tasks, and the second category of AI tasks includes one or more AI tasks other than the first category of AI tasks.

[0071] In the above method, the first type of task can be one or more tasks preset by the standard or system, such as CSI compression feedback, beam prediction, positioning enhancement, and transceiver data transmission. Users of the communication system may subsequently use the second type of AI tasks. However, considering that the AI ​​tasks involved in each terminal, network device, or each area are different, a general learning-related state is set for the second type of tasks, namely the public learning-related state, to improve the reporting efficiency of learning-related state information and save resources.

[0072] In another possible implementation, the first indication information is a task identifier.

[0073] In the above method, the first indication information may be used to indicate whether the learning-related status information comes from a specific task alone, or may be used in addition to indicating whether the learning-related status information comes from a specific task or a public task.

[0074] In another possible implementation, the environmental status information includes first parameter information, where the first parameter information includes one or more of a system parameter, a scene type, and a model identifier.

[0075] In the above method, the specific content of the environmental status information is reflected.

[0076] In another possible implementation, the environmental status information includes multiple task identifiers and multiple first parameter information, and each task identifier in the multiple task identifiers is used to represent an AI task associated with one first parameter information in the multiple first parameter information.

[0077] In the above method, the content of the environmental state information associated with multiple AI tasks is reflected.

[0078] In another possible implementation, the environmental status information is carried in MAC-CE signaling, and the MAC-CE signaling includes a task identification field, a system parameter field, a scenario type field, and a model identification field, the task identification field includes the task identification, the system parameter field includes the system parameters, the scenario type field includes the scenario type, and the model identification field includes the model identification.

[0079] In the above method, the field format of the MAC-CE signaling carrying the environmental status information is reflected.

[0080] In another possible implementation, the model performance information includes second parameter information, which includes one or more of a model identifier, model performance, fallback performance, and fallback indication. The model performance is used to characterize the performance of the model used for the associated AI task, and the fallback performance is used to characterize the system performance of the associated AI task after the model execution falls back to non-model execution. The fallback indication is used to characterize whether the associated AI task needs to be executed back, and the model identifier includes the model identifier of one or more models in the model associated with executing the AI ​​task.

[0081] In the above method, the specific content of the model performance information is reflected.

[0082] In another possible implementation, the model performance information includes multiple task identifiers and multiple second parameter information, and each task identifier in the multiple task identifiers is used to represent an AI task associated with a second parameter information in the multiple second parameter information.

[0083] In the above method, the content of model performance information related to multiple AI tasks is reflected.

[0084] In another possible embodiment, the model performance information is carried in MAC-CE signaling, and the MAC-CE signaling includes a task identification field, a model identification field, a model performance field, a fallback performance field and a fallback indication field, the task identification field includes the task identification, the model identification field includes the model identification, the model performance field includes the model performance, the fallback performance field includes the fallback performance, and the fallback indication field includes the fallback indication.

[0085] In the above method, the field format of the MAC-CE signaling that carries the model performance information is reflected.

[0086] In another possible embodiment, the data set status information includes third parameter information, and the third parameter information includes one or more of data set size, data collection time, data collection location and data distribution description, wherein the data collection time includes the collection time of the latest data sample in the data set collected for the associated AI task and the collection time of the oldest data sample in the data set, the data set size is used to characterize the number of data samples in the data set collected for the associated AI task, the data collection location is used to characterize the location of the data in the data set collected for the associated AI task, and the data distribution description is used to characterize the distribution of data in the data set collected for the associated AI task.

[0087] In the above method, the specific content of the data set status information is reflected.

[0088] In another possible implementation, the data set status information includes multiple task identifiers and multiple third parameter information, and each task identifier in the multiple task identifiers is used to represent the AI ​​task associated with one third parameter information in the multiple third parameter information.

[0089] In the above method, the content of the data set status information associated with multiple AI tasks is reflected.

[0090] In another possible implementation, the data set status information is carried in MAC-CE signaling, and the MAC-CE signaling includes a task identification field, a data set size field, a latest data acquisition time field, an oldest data acquisition time field, a data acquisition location field, and a data distribution description field. The task identification field includes the task identification, the data set size field includes the data set size, the latest data acquisition time field includes the latest data acquisition time, the oldest data acquisition time field includes the oldest data acquisition time, the data acquisition location field includes the data acquisition location, and the data distribution description field includes the data distribution description.

[0091] In the above method, the field format of the MAC-CE signaling carrying the data set status information is reflected.

[0092] In a third aspect, an embodiment of the present application provides a communication device, which includes a module or unit for implementing the method described in the first aspect or any possible implementation method of the first aspect.

[0093] It should be noted that the processor included in the communication device described in the third aspect above may be a processor specifically used to execute these methods (referred to as a dedicated processor for ease of distinction), or may be a processor that executes these methods by calling a computer program, such as a general-purpose processor. Optionally, at least one processor may include both a dedicated processor and a general-purpose processor.

[0094] Optionally, the computer program may be stored in a memory. For example, the memory may be a non-transitory memory, such as a read-only memory (ROM), which may be integrated with the processor on the same device or provided on separate devices. The embodiments of the present application do not limit the type of memory or the configuration of the memory and the processor.

[0095] In a possible implementation, the at least one memory is located outside the communication device.

[0096] In another possible implementation, the at least one memory is located within the communication device.

[0097] In another possible implementation, part of the at least one memory is located inside the communication device, and another part of the memory is located outside the communication device.

[0098] In this application, the processor and the memory may also be integrated into one device, that is, the processor and the memory may also be integrated together.

[0099] In a fourth aspect, an embodiment of the present application provides another communication device, which includes a module or unit for implementing the method described in the second aspect or any possible implementation method of the second aspect.

[0100] It should be noted that the processor included in the communication device described in the fourth aspect above may be a processor specifically used to execute these methods (referred to as a dedicated processor for ease of distinction), or may be a processor that executes these methods by calling a computer program, such as a general-purpose processor. Optionally, the at least one processor may include both a dedicated processor and a general-purpose processor.

[0101] Optionally, the computer program may be stored in a memory. For example, the memory may be a non-transitory memory, such as a read-only memory (ROM), which may be integrated with the processor on the same device or provided on separate devices. The embodiments of the present application do not limit the type of memory or the configuration of the memory and the processor.

[0102] In a possible implementation, the at least one memory is located outside the communication device.

[0103] In another possible implementation, the at least one memory is located within the communication device.

[0104] In another possible implementation, part of the at least one memory is located inside the communication device, and another part of the memory is located outside the communication device.

[0105] In this application, the processor and the memory may also be integrated into one device, that is, the processor and the memory may also be integrated together.

[0106] In a fifth aspect, an embodiment of the present application provides a communication device, which includes a processor; when the processor calls a computer program or instruction in a memory, it implements the method described in the first aspect or any one of the first aspects.

[0107] In a sixth aspect, an embodiment of the present application provides a communication device, which includes a processor; when the processor calls a computer program or instruction in a memory, it implements the method described in the aforementioned second aspect or any one of the second aspects.

[0108] In the seventh aspect, an embodiment of the present application provides a communication device, which includes a logic circuit and an interface, and the logic circuit and the interface are coupled; the interface is used to input and / or output information, and the logic circuit is used to execute the method described in any aspect of the first aspect to the second aspect or any possible implementation method.

[0109] In an eighth aspect, an embodiment of the present application provides a communication system, the communication system comprising the communication device as described in the third aspect and the communication device as described in the fourth aspect; or

[0110] The communication system includes the communication device as described in the fifth aspect and the communication device as described in the sixth aspect.

[0111] In a ninth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions. When the instructions are executed on at least one processor, the method described in any one of the first to second aspects above is implemented.

[0112] In a tenth aspect, the present application provides a computer program product, comprising computer instructions that, when executed on at least one processor, implement the method described in any one of the first and second aspects. The computer program product may be a software installation package. When the aforementioned method is required, the computer program product may be downloaded and executed on a computing device.

[0113] The beneficial effects of the technical methods provided in aspects 3 to 10 of this application can refer to the beneficial effects of the technical solutions in aspects 1 or 2, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0114] The following is a brief introduction to the drawings used in describing the embodiments.

[0115] FIG1 is a schematic diagram of a system architecture of a communication system 100 provided in an embodiment of the present application;

[0116] FIG2 is an interactive diagram of a communication method provided in an embodiment of the present application;

[0117] FIG3 is a schematic diagram of a process for periodically reporting learning-related status information provided by an embodiment of the present application;

[0118] FIG4 is a schematic diagram of a process for triggering reporting of learning-related status information provided by an embodiment of the present application;

[0119] FIG5 is a schematic diagram of a task identification provided in an embodiment of the present application;

[0120] FIG6 is a schematic structural diagram of a communication device 60 provided in an embodiment of the present application;

[0121] FIG7 is a schematic structural diagram of another communication device 70 provided in an embodiment of the present application;

[0122] FIG8 is a schematic structural diagram of a communication device 80 provided in an embodiment of the present application;

[0123] FIG9 is a schematic structural diagram of another communication device 90 provided in an embodiment of the present application. DETAILED DESCRIPTION

[0124] The embodiments of the present application are described below in conjunction with the drawings in the embodiments of the present application.

[0125] Please refer to Figure 1, which is a schematic diagram of the system architecture of a communication system 100 provided in an embodiment of the present application. As shown in Figure 1, the communication system 100 includes a radio access network (RAN) 110, a core network (CN) 120, and the Internet 130. RAN 110 includes at least one RAN node (such as 1101a and 1101b in Figure 1, collectively referred to as 1101) and at least one terminal (such as 1102a-1102j in Figure 1, collectively referred to as 1102). RAN 110 may also include other RAN nodes, such as wireless relay devices and / or wireless backhaul devices. Terminal 1102 is wirelessly connected to RAN node 1101. RAN node 110 is wirelessly or wiredly connected to core network 120. The core network devices in core network 120 and RAN node 1101 in RAN 110 can be different physical devices, or they can be the same physical device that integrates core network logical functions and radio access network logical functions.

[0126] The RAN 110 may be a cellular system related to the Third Generation Partnership Project (3GPP), such as a 4G or 5G mobile communication system, or a future-oriented evolutionary system (such as a 6G mobile communication system). The RAN 110 may also be an open access network (O-RAN or ORAN) or a cloud radio access network (CRAN). The RAN 110 may also be a communication system that integrates two or more of the above systems.

[0127] RAN node 1101, sometimes also referred to as access network equipment, RAN entity, or access node, constitutes part of the communication system and facilitates wireless access for terminals. Multiple RAN nodes 1101 in the communication system 100 can be of the same or different types. In some scenarios, the roles of RAN node 1101 and terminal 1102 are relative. For example, mobile device 1102i in Figure 1 can be a helicopter or drone, which can be configured as a mobile base station. For terminals 1102j accessing the RAN 110 via mobile device 1102i, mobile device 1102i is a base station; however, for base station 1101a, mobile device 1102i is a terminal. RAN node 1101 and terminal 1102 are sometimes referred to as communication devices. For example, 1101a and 1101b in Figure 1 can be understood as communication devices with base station functionality, and 1102a-1102j can be understood as communication devices with terminal functionality.

[0128] In one possible scenario, a RAN node may be a base station (BS), an evolved NodeB (eNodeB), an access point (AP), a transmission reception point (TRP), a next-generation NodeB (gNB), a next-generation NodeB in a sixth-generation (6G) mobile communication system, or a base station in a future mobile communication system. A RAN node may be a macro BS (such as 1101a in Figure 1 ), a micro BS or an indoor BS (such as 1101b in Figure 1 ), a relay node or a donor node, or a radio controller in a CRAN scenario. Alternatively, a RAN node may be a server, a wearable device, a vehicle, or an onboard device. For example, the access network device in vehicle-to-everything (V2X) technology may be a roadside unit (RSU).

[0129] In another possible scenario, multiple RAN nodes collaborate to assist the terminal in achieving wireless access, and different RAN nodes respectively implement part of the functions of the base station. For example, the RAN node can be a centralized unit (CU), a distributed unit (DU), a CU-control plane (CP), a CU-user plane (UP), or a radio unit (RU). The CU and DU can be set separately, or they can be included in the same network element, such as a baseband unit (BBU). The RU can be included in a radio frequency device or radio frequency unit, such as a remote radio unit (RRU), an active antenna unit (AAU), or a remote radio head (RRH).

[0130] In different systems, CU (or CU-CP and CU-UP), DU or RU may also have different names, but those skilled in the art can understand their meanings. For example, in the ORAN system, CU may also be called O-CU (Open CU), DU may also be called O-DU, CU-CP may also be called O-CU-CP, CU-UP may also be called O-CU-UP, and RU may also be called O-RU. For the convenience of description, this application uses CU, CU-CP, CU-UP, DU and RU as examples for description. Any unit of CU (or CU-CP, CU-UP), DU and RU in this application can be implemented by a software module, a hardware module, or a combination of a software module and a hardware module.

[0131] A terminal may also be referred to as a terminal device, user equipment (UE), mobile station, mobile terminal, etc. The terminal can be widely used in various scenarios, such as device-to-device (D2D), vehicle-to-everything (V2X) communication, machine-type communication (MTC), Internet of Things (IOT), virtual reality, augmented reality, industrial control, autonomous driving, telemedicine, smart grid, smart furniture, smart office, smart wearable, smart transportation, smart city, etc. The terminal can be a mobile phone, tablet computer, computer with wireless transceiver function, wearable device, vehicle, drone, helicopter, airplane, ship, robot, robotic arm, smart home device, etc. The embodiments of the present application do not limit the device form of the terminal.

[0132] In the embodiments of the present application, the network device mentioned later may be the aforementioned RAN node 1101, the core network 120, or the mobile device 1102i, and this application does not impose any further limitations on this. The terminal dynamically sends learning-related status information associated with the AI ​​task to the network device to support the network device in performing operations such as lifecycle management and learning strategy adjustment on the model associated with the AI ​​task.

[0133] Before introducing the solution of this application, the following points are explained.

[0134] (1) In this application, “indication” may include direct indication, indirect indication, explicit indication, and implicit indication. When describing that a certain indication information is used to indicate A, it can be understood that the indication information carries A, directly indicates A, or indirectly indicates A.

[0135] In this application, the information indicated by the indication information is referred to as the information to be indicated. In the specific implementation process, there are many ways to indicate the information to be indicated, such as but not limited to, the information to be indicated can be directly indicated, such as the information to be indicated itself or the index of the information to be indicated. The information to be indicated can also be indirectly indicated by indicating other information, wherein there is an association between the other information and the information to be indicated. It is also possible to indicate only a part of the information to be indicated, while the other parts of the information to be indicated are known or agreed in advance. For example, the indication of specific information can also be achieved with the help of the arrangement order of each information agreed in advance (for example, stipulated by the protocol), thereby reducing the indication overhead to a certain extent. In addition, the information to be indicated can be sent together as a whole, or it can be divided into multiple sub-information and sent separately, and the sending period and / or sending time of these sub-information can be the same or different.

[0136] (2) In this application, "sending" and "receiving" indicate the direction of signal transmission. For example, "sending information to XX" can be understood as the destination of the information being XX, which can include direct sending through the air interface, and also include indirect sending through the air interface by other units or modules. "Receiving information from YY" can be understood as the source of the information being YY, which can include direct receiving from YY through the air interface, and also include indirect receiving from YY through the air interface from other units or modules. "Sending" can also be understood as the "output" of the chip interface, and "receiving" can also be understood as the "input" of the chip interface. In other words, sending and receiving can be carried out between devices, for example, between a network device and a terminal device, or can be carried out within a device, for example, sending or receiving between components, modules, chips, software modules or hardware modules within the device through a bus, a line or an interface.

[0137] Please refer to Figure 2, which is an interactive diagram of a communication method provided in an embodiment of the present application. The communication method is described from the perspective of interaction between a terminal and a network device. Optionally, the communication method can be applied to the aforementioned system, such as the communication system shown in Figure 1.

[0138] The communication method includes but is not limited to steps S201 to S204 as shown below. It should be understood that for the convenience of description, the description is given in the order of steps S201 to S204, and is not intended to limit the execution to the above order. The embodiment of the present application does not limit the order of execution, execution time, number of executions, etc. of the above one or more steps, and other steps can be interspersed between or before and after these steps as needed. Among them, steps S201 to S204 are specifically as follows:

[0139] Step S201: The terminal obtains learning-related status information associated with the AI ​​task.

[0140] Among them, learning-related status information includes one or more of resource status information, environment status information, model performance information and data set status information. Learning-related status information is used by network devices to perform lifecycle management of models associated with AI tasks (wherein, multiple models may be deployed under one AI task, and network devices can perform lifecycle management of some or all models).

[0141] In an embodiment of the present application, before obtaining the learning-related status information associated with the AI ​​task, the terminal may obtain a sending method of the learning-related status information (for example, a condition for sending the learning-related status information) by receiving instruction information sent by the network device. Therefore, optionally, the terminal obtains a sending method of the learning-related status information associated with the AI ​​task, specifically as follows:

[0142] In the first mode, the indication information includes a reference time; when the reference time is met, the terminal sends learning-related status information to the network device.

[0143] Specifically, please refer to Figure 3, which is a flow chart of a periodic reporting of learning-related status information provided by an embodiment of the present application. As shown in Figure 3, for example, the reference time is 30 minutes, and the network device can send the set duration of the timer to the terminal, and then the terminal starts the timer. When the set duration of the timer reaches 30 minutes, the terminal obtains the learning-related status information of the associated AI task and then sends it to the network device.

[0144] In the second method, the indication information includes a trigger condition; when the trigger condition is met, the terminal sends learning-related status information to the network device.

[0145] Specifically, please refer to Figure 4, which is a flow chart of a triggered reporting of learning-related status information provided by an embodiment of the present application. As shown in Figure 4, for example, the trigger condition may be that the reasoning accuracy of the model is lower than a preset threshold. If the associated AI task is "beam prediction", the reasoning accuracy of the model associated with the task is 80%, which is lower than the preset threshold of 90%, then the trigger condition for reporting the learning-related status information is met. At this time, the terminal can obtain the learning-related status information of the associated AI task and then send it to the network device.

[0146] Optionally, the terminal sends instruction information including a task identifier to the network device.

[0147] Among them, the indication information of the task identifier is used to indicate the associated AI task. The task identifier can be carried in the learning-related state information, or it can exist independently without being carried in the learning-related state information. This application does not limit this. For example, please refer to Figure 5. Figure 5 is a schematic diagram of a task identifier provided by an embodiment of the present application. As shown in Figure 5, if the associated AI task is "beam prediction", Figure 5 (a) shows that the task identifier is carried in the learning-related state information, that is, each signaling carried by the resource state information, the environment state information, the model performance information and the data set state information includes the task identifier separately. Figure 5 (b) shows that the task identifier is a total task identifier that exists independently without being carried in the learning-related state information. To facilitate the description of subsequent embodiments, this application takes the case where the task identifier is carried in the learning-related state information as an example for explanation, and will not be repeated in detail later.

[0148] It should be noted that the timer duration information and trigger condition information sent by the network device both carry task identification information, allowing the terminal to understand which specific task the current timer and trigger condition are associated with. Accordingly, the terminal also needs to carry the task identification in the learning-related status information it reports back to indicate which task the reported learning-related status information is associated with.

[0149] Next, the signaling reporting design of each information in the learning-related status information is elaborated in detail.

[0150] In an exemplary design, the resource status information includes a resource margin, and the resource margin includes a computing power margin and / or a storage margin. Taking the resource margin as an example, if the resource status information is carried in the signaling of the media access control-control element MAC-CE, the signaling of the MAC-CE includes a task identification field (i.e., the "TaskID" field in Table 1) and a CPH field (i.e., the "CPH" field in Table 1), the task identification field includes a task identifier, and the CPH field includes the resource margin.

[0151] Optionally, the "R" field includes a MAC-CE reserved bit.

[0152] The "G" field indicates whether the reported status is a public learning-related status or a specific task learning-related status. For example, "1" indicates that the signaling reports a public learning-related status, and "0" indicates that the signaling reports a specific task learning-related status.

[0153] The above fields are shown in Table 1:

[0154] Table 1

[0155] Optionally, the terminal sends first indication information to the network device.

[0156] Among them, the first indication information is used to indicate that the learning-related status information comes from the first category of AI tasks or the second category of AI tasks. The first category of AI tasks includes one or more preset tasks, and the second category of AI tasks includes one or more AI tasks other than the first category of AI tasks. It should be noted that the first category of AI tasks may include specific tasks corresponding to the specific task learning-related status in the above-mentioned "G" field, and the second category of AI tasks may include other tasks corresponding to the public learning-related status in the above-mentioned "G" field. For example, the first category of tasks may be one or more tasks preset by the standard or system, such as one or more of CSI compression feedback, beam prediction, positioning enhancement, and transceiver data transmission. Users using the communication system may subsequently use the second category of AI tasks. However, considering that the AI ​​tasks involved in each terminal, network device, or each area are different, a general learning-related status is set for the second category of tasks, namely the public learning-related status mentioned in Table 1, to improve the reporting efficiency of learning-related status information and save resources.

[0157] Further optionally, the first indication information is a task identifier.

[0158] That is, the "G" field can reuse the "TaskID" field, as shown in Table 1 above; or it can be used alone to indicate the associated AI task, as shown in Table 2:

[0159] Table 2

[0160] In addition, it should be noted that margin refers to the resources currently available in the space. For example, if the total storage capacity of the computing power margin is 100G and the existing AI tasks have occupied 60G, the remaining 40G is the margin. The resource margin can be indicated by a level index, that is, the resource margin value is divided into multiple levels, and each level defines its index. For example, if the resource margin value is N, it is evenly divided into 0, 1, 2, ..., N levels, and the corresponding level indexes CPHLevel are Headroom_0, Headroom_1, Headroom_2, ..., Headroom_N, respectively, as shown in Table 3:

[0161] Table 3

[0162] When resource status information of multiple tasks needs to be reported, the resource status information includes multiple task identifiers and multiple resource margins. Each of the multiple task identifiers is used to represent an AI task associated with one of the multiple resource margins. (It should be noted that the multiple task identifiers here correspond to multiple tasks, which is different from the associated AI tasks described in step S201. The associated AI tasks in step S201 can be one or more tasks, while the multiple tasks here only refer to the existence of resource status information of multiple tasks that needs to be reported. The subsequent embodiments are the same as the explanation here, so they are not repeated here.) For example, the following signaling can be used, as shown in Table 4:

[0163] Table 4

[0164] ...

[0165] Optional, "C n The " field is used to indicate whether the learning-related status information related to task N appears in this signaling. For example, if C1 is 1, it means that the parameters related to Task ID1 appear in this signaling, or it can be 0, which means that the parameters related to Task ID1 appear in this signaling. The embodiment of the present application does not make too many restrictions on this. For other fields, please refer to the corresponding explanations of the fields in Table 1 above, which will not be repeated here.

[0166] In an exemplary design, the environmental status information includes first parameter information, and the first parameter information includes one or more of a system parameter, a scenario type, and a model identifier. Taking the environmental status information carried in MAC-CE signaling as an example, the MAC-CE signaling includes a task identification field (i.e., the "TaskID" field in Table 5), a system parameter field (i.e., the "Systemparameter" field in Table 5), a scenario type field (i.e., the "Scenariotype" field in Table 5), and a model identification field (i.e., the "Pairingcondition" field in Table 5), the task identification field includes a task identifier, the system parameter field includes system parameters, the scenario type field includes a scenario type, and the model identification field includes a model identifier.

[0167] As shown in Table 5:

[0168] Table 5

[0169] Optionally, the "R" field includes a MAC-CE reserved bit.

[0170] The "System parameter" field includes the parameter configuration associated with the task or the change of the system parameters. For different tasks, the relevant system parameters will also be different. Taking the four tasks in the first category, such as Channel State Information (CSI) compression feedback, beam prediction, positioning enhancement, and transceiver data transmission, as an example, the possible relevant system parameters of these tasks are shown in Table 6:

[0171] Table 6

[0172] The "Scenariotype" field contains the scenario in which the task is performed. The scenario type can include one or more of indoor or outdoor, urban or suburban, and line-of-sight (LOS) and non-line-of-sight (NLOS) transmissions of wireless signals. For example, the scenario for a task can switch from an indoor location in an urban area to an outdoor location in a suburban area, or vice versa. This application does not impose any specific restrictions on this.

[0173] The "PairingCondition" field contains the model identifier of the model used to perform the task and match the communication peer. For example, the model identifier is used to ensure that the model used by the terminal and network device associated with the "CSI Compression Feedback" task remains consistent, so that the model can be used to properly collect environmental status information corresponding to the model associated with the "CSI Compression Feedback" task.

[0174] When there are multiple tasks whose environmental status information needs to be reported, the environmental status information includes multiple task identifiers and multiple first parameter information. Each of the multiple task identifiers is used to represent the AI ​​task associated with one of the multiple first parameter information. For example, the following signaling can be used, as shown in Table 7:

[0175] Table 7

[0176] ...

[0177] In an exemplary design, the model performance information includes second parameter information, and the second parameter information includes one or more of a model identifier, a model performance, a fallback performance, and a fallback indication.

[0178] Among them, model performance is used to characterize the performance of the model used for the associated AI task; fallback performance is used to characterize the system performance after the associated AI task falls back from model execution to non-model execution; fallback indication is used to characterize whether the associated AI task needs to be rolled back, for example, if this field is "1", it means that the associated AI task needs to be rolled back, and if it is "0", it means that rollback is not required; the model identifier includes the model identifier of one or more models in the model associated with the execution of the AI ​​task.

[0179] Taking the model performance information carried in the MAC-CE signaling as an example, the MAC-CE signaling includes a task identification field (i.e., the "TaskID" field in Table 8), a model identification field (i.e., the "ModelID" field in Table 8), a model performance field (i.e., the "Modelperformance" field in Table 8), a fallback performance field (i.e., the "Fallback performance" field in Table 8), and a fallback indication field (i.e., the "F" field in Table 8). The task identification field includes the task identification, the model identification field includes the model identification, the model performance field includes the model performance, the fallback performance field includes the fallback performance, and the fallback indication field includes the fallback indication.

[0180] As shown in Table 8:

[0181] Table 8

[0182] ...

[0183] Optionally, the "R" field includes a MAC-CE reserved bit.

[0184] “M nThe " field is used to indicate whether the performance of the Nth model that completes a specific task appears in this signaling. For example, if M1 is 1, it indicates that the performance of the first model that completes the specific task "CSI compression feedback" appears in this signaling, or M1 can be 0, which indicates that the performance of the first model that completes the specific task "CSI compression feedback" appears in this signaling. The embodiment of the present application does not impose too many restrictions on this. For other fields, please refer to the corresponding explanations of the fields in Table 8 above, which will not be repeated here.

[0185] When model performance information of multiple tasks needs to be reported, the model performance information includes multiple task identifiers and multiple second parameter information. Each task identifier in the multiple task identifiers is used to represent the AI ​​task associated with one second parameter information in the multiple second parameter information, as shown in Table 9:

[0186] Table 9

[0187] ......

[0188] ......

[0189] ......

[0190] In an exemplary design, the data set status information includes third parameter information, and the third parameter information includes one or more of the data set size, data collection time, data collection location, and data distribution description.

[0191] Among them, the data collection time includes the collection time of the latest data sample and the collection time of the oldest data sample in the dataset collected for the associated AI task. The dataset size is used to characterize the number of data samples in the dataset collected for the associated AI task. The data collection location is used to characterize the location of the data in the dataset collected for the associated AI task. The data distribution description is used to characterize the distribution of data in the dataset collected for the associated AI task.

[0192] Taking the data set status information carried in the MAC-CE signaling as an example, the MAC-CE signaling includes a task identification field (i.e., the "TaskID" field in Table 10), a data set size field (i.e., the "DatasetSize" field in Table 10), the latest data collection time field (i.e., the "Latestdatatimestamp" field in Table 10), the oldest data collection time field (i.e., the "Oldest data time stamp" field in Table 10), the data collection position field (i.e., the "Collectionposition" field in Table 10) and the data distribution description field (i.e., the "Distributiondescription" field in Table 10), the task identification field includes the task identification, the data set size field includes the data set size, the latest data collection time field includes the latest data collection time, the oldest data collection time field includes the oldest data collection time, the data collection position field includes the data collection position, and the data distribution description field includes the data distribution description.

[0193] As shown in Table 10:

[0194] Table 10

[0195] Optionally, the "R" field includes a MAC-CE reserved bit.

[0196] When there are multiple task dataset status information that need to be reported, the dataset status information includes multiple task identifiers and multiple third parameter information. Each of the multiple task identifiers is used to represent the AI ​​task associated with one of the multiple third parameter information. For example, the following signaling can be used, as shown in Table 11:

[0197] Table 11

[0198] ......

[0199] Step S202: The terminal sends learning-related status information to the network device.

[0200] Step S203: The network device receives the learning-related status information of the associated AI task sent by the terminal.

[0201] Step S204: The network device performs lifecycle management on the model associated with the AI ​​task based on the learning-related status information.

[0202] Among them, lifecycle management includes one or more of model selection, model training, model reasoning, model transfer, model enabling, model switching, model rollback, model monitoring, etc. Specifically, the network device will manage the models associated with the AI ​​task based on the learning-related status information reported by the terminal. For example, the learning-related status reported by the terminal shows that the performance of the model associated with the AI ​​task has deteriorated. The network device may send a new model to the terminal, or there are already multiple models deployed on the terminal. The network device can instruct the terminal to switch from the current model with deteriorated performance to another model. The worst case is that the entire AI method is not suitable and can fall back to the traditional method. The above operations are used to take into account all models associated with the AI ​​task globally. In an embodiment of the present application, the efficiency and accuracy of lifecycle management of models associated with AI tasks can be improved by performing operations such as reporting learning-related status information for specific tasks.

[0203] The above describes in detail the method of the embodiment of the present application. The following provides an apparatus of the embodiment of the present application.

[0204] It can be understood that the multiple devices provided in the embodiments of the present application, such as communication devices, include hardware structures, software modules, or a combination of hardware structures and software structures corresponding to executing each function in order to implement the functions in the above method embodiments.

[0205] Those skilled in the art should easily appreciate that, in combination with the units and algorithm steps of each example described in the embodiments disclosed herein, the embodiments of the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different device implementations to implement the aforementioned method embodiments in different usage scenarios, and different implementations of the devices should not be considered to exceed the scope of the embodiments of the present application.

[0206] The embodiments of the present application may divide the device into functional modules. For example, each functional module may be divided according to each function, or two or more functions may be integrated into one functional module. The above-mentioned integrated modules may be implemented in the form of hardware or software functional modules. It should be noted that the division of modules in the embodiments of the present application is schematic and is only a logical functional division. In actual implementation, other division methods may be used.

[0207] For example, in the case of dividing the functional modules of the device in an integrated manner, the present application cites several possible communication devices.

[0208] Please refer to Figure 6, which is a schematic diagram of the structure of a communication device 60 provided in an embodiment of the present application. The communication device 60 may be the terminal shown in Figure 1 or a device in the terminal, such as a chip, software module, integrated circuit, etc. The communication device 60 includes a communication unit 601 and a processing unit 602, wherein:

[0209] The communication unit 601 is a general term for devices capable of establishing a communication connection and sending and receiving data in the embodiments of the present application. The communication device 60 may also be other communication units. For example, the communication unit 601 may be an input / output interface, or may be called a communication interface, or an interface circuit, or an interface, etc. Alternatively, the communication unit 601 may be a sending unit and a receiving unit, the sending unit may be an output interface, the receiving unit may be an input interface, and the sending unit and the receiving unit may be integrated into one unit, such as an input / output interface. The communication device 60 is used to implement the aforementioned communication method in the method embodiments of the present application, such as the communication method described in FIG.

[0210] The processing unit 602 may be one or more logic circuits, that is, the processing unit 602 may be implemented using logic circuits, and the communication unit 601 may be implemented using interfaces. The logic circuit may be a chip, a processing circuit, an integrated circuit, or a system-on-chip (SoC) chip, and the interface may be a communication interface, an input / output interface, a pin, or the like.

[0211] In the embodiments of the present application, the logic circuit and the interface may also be coupled to each other. The embodiments of the present application do not limit the specific connection method of the logic circuit and the interface. For example, the logic circuit may be used to execute the functions or steps implemented by the processing unit 602, and the interface may be used to execute the functions or steps implemented by the communication unit 601.

[0212] The communication device 60 is capable of implementing the functions of the terminal described in the embodiments of the present application. For example, the communication device 60 includes modules, units, or means corresponding to the steps involved in the terminal described in the embodiments of the present application. The functions, units, or means can be implemented by software, or by hardware, or by hardware executing corresponding software implementations, or by a combination of software and hardware. For details, please refer to the corresponding description in the aforementioned corresponding method embodiment.

[0213] In one possible design, the processing unit 602 is configured to obtain learning-related state information associated with the artificial intelligence (AI) task, wherein the learning-related state information includes one or more of resource state information, environment state information, model performance information, and dataset state information;

[0214] The communication unit 601 is used to send the learning-related status information, wherein the learning-related status information is used to perform lifecycle management on the model associated with the AI ​​task.

[0215] In an optional implementation, the communication unit 601 is further configured to:

[0216] Receive indication information, wherein the indication information is used to indicate a sending method of the learning-related status information.

[0217] In another optional implementation, the indication information includes a reference time;

[0218] In terms of sending the learning-related status information, the communication unit 601 is further configured to:

[0219] When the reference time is met, the learning-related status information is sent.

[0220] In another optional implementation, the indication information includes a trigger condition;

[0221] In terms of sending the learning-related status information, the communication unit 601 is further configured to:

[0222] When the trigger condition is met, the learning-related status information is sent.

[0223] In another optional implementation, the communication unit 601 is further configured to:

[0224] Sending indication information of a task identifier, wherein the indication information of the task identifier is used to indicate the associated AI task.

[0225] In another optional implementation, the resource status information includes resource margin, and the resource margin includes computing power margin and / or storage margin.

[0226] In another optional implementation, the resource status information includes multiple task identifiers and multiple resource margins, and each task identifier in the multiple task identifiers is used to represent an AI task associated with a resource margin in the multiple resource margins.

[0227] In another optional embodiment, the resource status information is carried in the signaling of the media access control-control element MAC-CE, and the signaling of the MAC-CE includes a task identification field and a resource margin CPH field, the task identification field includes the task identification, and the CPH field includes the resource margin.

[0228] In another optional implementation, the communication unit 601 is further configured to:

[0229] Sending first indication information, wherein the first indication information is used to indicate that the learning-related state information comes from a first category of AI tasks or a second category of AI tasks, the first category of AI tasks includes one or more preset tasks, and the second category of AI tasks includes one or more AI tasks other than the first category of AI tasks.

[0230] In another optional implementation, the first indication information is a task identifier.

[0231] In another optional implementation, the environmental status information includes first parameter information, and the first parameter information includes one or more of a system parameter, a scene type, and a model identifier.

[0232] In another optional embodiment, the environmental status information includes multiple task identifiers and multiple first parameter information, and each task identifier in the multiple task identifiers is used to represent the AI ​​task associated with one first parameter information in the multiple first parameter information.

[0233] In another optional embodiment, the environmental status information is carried in the MAC-CE signaling, and the MAC-CE signaling includes a task identification field, a system parameter field, a scenario type field and a model identification field, the task identification field includes the task identification, the system parameter field includes the system parameters, the scenario type field includes the scenario type, and the model identification field includes the model identification.

[0234] In another optional embodiment, the model performance information includes second parameter information, and the second parameter information includes one or more of a model identifier, a model performance, a fallback performance, and a fallback indication, wherein the model performance is used to characterize the performance of the model used for the associated AI task, the fallback performance is used to characterize the system performance of the associated AI task after the model execution falls back to the non-model execution, the fallback indication is used to characterize whether the associated AI task needs to be executed back, and the model identifier includes the model identifier of one or more models in the model associated with executing the AI ​​task.

[0235] In another optional embodiment, the model performance information includes multiple task identifiers and multiple second parameter information, and each task identifier in the multiple task identifiers is used to represent the AI ​​task associated with one second parameter information in the multiple second parameter information.

[0236] In another optional embodiment, the model performance information is carried in the MAC-CE signaling, and the MAC-CE signaling includes a task identification field, a model identification field, a model performance field, a fallback performance field and a fallback indication field, the task identification field includes the task identification, the model identification field includes the model identification, the model performance field includes the model performance, the fallback performance field includes the fallback performance, and the fallback indication field includes the fallback indication.

[0237] In another optional embodiment, the data set status information includes third parameter information, and the third parameter information includes one or more of the data set size, data collection time, data collection location and data distribution description, wherein the data collection time includes the collection time of the latest data sample in the data set collected for the associated AI task and the collection time of the oldest data sample in the data set, the data set size is used to characterize the number of data samples in the data set collected for the associated AI task, the data collection location is used to characterize the location of the data in the data set collected for the associated AI task, and the data distribution description is used to characterize the distribution of data in the data set collected for the associated AI task.

[0238] In another optional implementation, the data set status information includes multiple task identifiers and multiple third parameter information, and each task identifier in the multiple task identifiers is used to represent the AI ​​task associated with one third parameter information in the multiple third parameter information.

[0239] In another optional implementation, the data set status information is carried in MAC-CE signaling, and the MAC-CE signaling includes a task identification field, a data set size field, a latest data acquisition time field, a oldest data acquisition time field, a data acquisition location field and a data distribution description field, the task identification field includes the task identification, the data set size field includes the data set size, the latest data acquisition time field includes the latest data acquisition time, the oldest data acquisition time field includes the oldest data acquisition time, the data acquisition location field includes the data acquisition location, and the data distribution description field includes the data distribution description.

[0240] The embodiments of the present application and the above-mentioned method embodiments are based on the same concept, and the technical effects they bring are also the same. For the specific principles, please refer to the description of the above-mentioned embodiments, which will not be repeated here.

[0241] Please refer to Figure 7, which is a schematic diagram of the structure of another communication device 70 provided in an embodiment of the present application. The communication device 70 may be the network device shown in Figure 1 or a component in the network device, such as a chip, software module, integrated circuit, etc. The communication device 70 includes a communication unit 701 and a processing unit 702, wherein:

[0242] The communication unit 701 is a general term for devices capable of establishing a communication connection and sending and receiving data in the embodiments of the present application. The communication device 70 may also be other communication units. For example, the communication unit 701 may be an input / output interface, or may be called a communication interface, or an interface circuit, or an interface, etc. Alternatively, the communication unit 701 may be a sending unit and a receiving unit. The sending unit may be an output interface, and the receiving unit may be an input interface. The sending unit and the receiving unit are integrated into one unit, such as an input / output interface. The communication device 70 is used to implement the aforementioned communication method in the method embodiments of the present application, such as the communication method described in FIG.

[0243] The processing unit 702 may be one or more logic circuits, that is, the processing unit 702 may be implemented using logic circuits, and the communication unit 701 may be implemented using interfaces. The logic circuit may be a chip, a processing circuit, an integrated circuit, or a system-on-chip (SoC) chip, and the interface may be a communication interface, an input / output interface, a pin, or the like.

[0244] In the embodiments of the present application, the logic circuit and the interface may also be coupled to each other. The embodiments of the present application do not limit the specific connection method of the logic circuit and the interface. For example, the logic circuit may be used to execute the functions or steps implemented by the processing unit 702, and the interface may be used to execute the functions or steps implemented by the communication unit 701.

[0245] The communication device 70 is capable of implementing the functions of the terminal described in the embodiments of the present application. For example, the communication device 70 includes modules, units, or means corresponding to the steps involved in the terminal described in the embodiments of the present application. The functions, units, or means can be implemented by software, or by hardware, or by hardware executing corresponding software implementations, or by a combination of software and hardware. For details, please refer to the corresponding description in the aforementioned corresponding method embodiment.

[0246] In one possible design, the communication unit 701 is configured to receive learning-related status information associated with the AI ​​task, wherein the learning-related status information includes one or more of resource status information, environment status information, model performance information, and dataset status information;

[0247] The processing unit 702 is used to perform lifecycle management on the model associated with the AI ​​task based on the learning-related state information.

[0248] In an optional implementation, the communication unit 701 is further configured to:

[0249] Sending instruction information, wherein the instruction information is used to indicate a sending method of the learning-related status information.

[0250] In another optional implementation, the indication information includes a reference time, and the indication information is used to indicate that the learning-related status information should be sent when the reference time is met.

[0251] In another optional implementation, the indication information includes a trigger condition, and the indication information is used to indicate that the learning-related status information is sent when the trigger condition is met.

[0252] In another optional implementation, the communication unit 701 is further configured to:

[0253] Receive indication information of a task identifier, wherein the indication information of the task identifier is used to indicate the associated AI task.

[0254] In another optional implementation, the resource status information includes resource margin, and the resource margin includes computing power margin and / or storage margin.

[0255] In another optional implementation, the resource status information includes multiple task identifiers and multiple resource margins, and each task identifier in the multiple task identifiers is used to represent an AI task associated with a resource margin in the multiple resource margins.

[0256] In another optional implementation, the resource status information is carried in MAC-CE signaling, and the MAC-CE signaling includes a task identification field and a CPH field, the task identification field includes the task identification, and the CPH field includes the resource margin.

[0257] In another optional implementation, the communication unit 701 is further configured to:

[0258] Receive first indication information, wherein the first indication information is used to indicate that the learning-related state information comes from a first category of AI tasks or a second category of AI tasks, the first category of AI tasks includes one or more preset tasks, and the second category of AI tasks includes one or more AI tasks other than the first category of AI tasks.

[0259] In another optional implementation, the first indication information is a task identifier.

[0260] In another optional implementation, the environmental status information includes first parameter information, and the first parameter information includes one or more of a system parameter, a scene type, and a model identifier.

[0261] In another optional embodiment, the environmental status information includes multiple task identifiers and multiple first parameter information, and each task identifier in the multiple task identifiers is used to represent the AI ​​task associated with one first parameter information in the multiple first parameter information.

[0262] In another optional embodiment, the environmental status information is carried in the MAC-CE signaling, and the MAC-CE signaling includes a task identification field, a system parameter field, a scenario type field and a model identification field, the task identification field includes the task identification, the system parameter field includes the system parameters, the scenario type field includes the scenario type, and the model identification field includes the model identification.

[0263] In another optional embodiment, the model performance information includes second parameter information, and the second parameter information includes one or more of a model identifier, a model performance, a fallback performance, and a fallback indication, wherein the model performance is used to characterize the performance of the model used for the associated AI task, the fallback performance is used to characterize the system performance of the associated AI task after the model execution falls back to the non-model execution, the fallback indication is used to characterize whether the associated AI task needs to be executed back, and the model identifier includes the model identifier of one or more models in the model associated with executing the AI ​​task.

[0264] In another optional embodiment, the model performance information includes multiple task identifiers and multiple second parameter information, and each task identifier in the multiple task identifiers is used to represent the AI ​​task associated with one second parameter information in the multiple second parameter information.

[0265] In another optional embodiment, the model performance information is carried in the MAC-CE signaling, and the MAC-CE signaling includes a task identification field, a model identification field, a model performance field, a fallback performance field and a fallback indication field, the task identification field includes the task identification, the model identification field includes the model identification, the model performance field includes the model performance, the fallback performance field includes the fallback performance, and the fallback indication field includes the fallback indication.

[0266] In another optional embodiment, the data set status information includes third parameter information, and the third parameter information includes one or more of the data set size, data collection time, data collection location and data distribution description, wherein the data collection time includes the collection time of the latest data sample in the data set collected for the associated AI task and the collection time of the oldest data sample in the data set, the data set size is used to characterize the number of data samples in the data set collected for the associated AI task, the data collection location is used to characterize the location of the data in the data set collected for the associated AI task, and the data distribution description is used to characterize the distribution of data in the data set collected for the associated AI task.

[0267] In another optional implementation, the data set status information includes multiple task identifiers and multiple third parameter information, and each task identifier in the multiple task identifiers is used to represent the AI ​​task associated with one third parameter information in the multiple third parameter information.

[0268] In another optional implementation, the data set status information is carried in MAC-CE signaling, and the MAC-CE signaling includes a task identification field, a data set size field, a latest data acquisition time field, a oldest data acquisition time field, a data acquisition location field and a data distribution description field, the task identification field includes the task identification, the data set size field includes the data set size, the latest data acquisition time field includes the latest data acquisition time, the oldest data acquisition time field includes the oldest data acquisition time, the data acquisition location field includes the data acquisition location, and the data distribution description field includes the data distribution description.

[0269] The embodiments of the present application and the above-mentioned method embodiments are based on the same concept, and the technical effects they bring are also the same. For the specific principles, please refer to the description of the above-mentioned embodiments, which will not be repeated here.

[0270] In the various embodiments of the present application, unless otherwise specified or there is a logical conflict, the terms and / or descriptions between the various embodiments are consistent and can be referenced by each other. The technical features in different embodiments can be combined to form new embodiments according to their inherent logical relationships.

[0271] Please refer to Figure 8, which is a schematic diagram of the structure of a communication device 80 provided in an embodiment of the present application. The communication device 80 can be a terminal or a component within the terminal (such as a chip, a software module, or a hardware module). The communication device 80 may include at least one processor 801. Optionally, the communication device 80 may also include a communication interface 802. Further optionally, the communication device 80 may also include at least one memory 803. Further optionally, the communication device 80 may also include a bus 804, wherein the processor 801, the communication interface 802, and the memory 803 are connected via the bus 804.

[0272] Optionally, the processor 801 may be a processor specifically used to execute these methods (referred to as a dedicated processor for ease of distinction), or may be a processor that executes these methods by calling a computer program, such as a general-purpose processor. Optionally, the at least one processor may include both a dedicated processor and a general-purpose processor. Optionally, when the communication device 80 includes at least one processor 801, the computer program may be stored in the memory 803.

[0273] The communication interface 802 can be used to provide information input or output for the at least one processor. Furthermore, the communication interface 802 can be used to receive data sent externally and / or send data externally. The communication interface 802 can be a wired link interface such as an Ethernet cable, or a wireless link interface (wireless fidelity (WiFi), Bluetooth, general wireless transmission, in-vehicle short-range communication technology, or other short-range wireless communication technology). Optionally, the communication interface 802 can also include a transmitter (such as a radio frequency transmitter, antenna, etc.) or a receiver coupled to the interface.

[0274] Memory 803 is used to provide storage space for storing data such as the operating system and computer programs. Memory 803 can be one or a combination of random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or compact disc read-only memory (CD-ROM).

[0275] The at least one processor 801 in the communication device 80 is configured to execute the aforementioned communication method, such as the steps executed by the terminal in the embodiment shown in Figure 2. For related details, please refer to the above description and will not be repeated here.

[0276] Please refer to Figure 9, which is a schematic diagram of the structure of another communication device 90 provided in an embodiment of the present application. The communication device 90 can be a network device or a component within the network device (such as a chip, a software module, or a hardware module). The communication device 90 may include at least one processor 901. Optionally, the communication device 90 may also include a communication interface 902. Further optionally, the communication device 90 may also include at least one memory 903. Further optionally, the communication device 90 may also include a bus 904, wherein the processor 901, the communication interface 902, and the memory 903 are connected via the bus 904.

[0277] It should be understood that the division of the various units in the communication devices of Figures 8 and 9 is merely a division of logical functions. In actual implementation, they can be fully or partially integrated into a physical entity, or they can be physically separated. In addition, the units in the communication device can be implemented in the form of a processor calling software; for example, the communication device includes a processor, the processor is connected to a memory, and the memory stores instructions. The processor calls the instructions stored in the memory to implement any of the above methods or realize the functions of the various units of the communication device, wherein the processor is, for example, a general-purpose processor, such as a central processing unit (CPU) or a microprocessor, and the memory is a memory within the communication device or a memory outside the communication device. Alternatively, the units in the communication device can be implemented in the form of hardware circuits, and the functions of some or all of the units can be realized by designing the hardware circuits. The hardware circuit can be understood as one or more processors. For example, in one implementation, the hardware circuit is an application-specific integrated circuit (ASIC), which realizes the functions of some or all of the above units by designing the logical relationship of the components in the circuit. For another example, in another implementation, the hardware circuit can be implemented by a programmable logic device (PLD). Taking a field programmable gate array (FPGA) as an example, it can include a large number of logic gate circuits, and the connection relationship between the logic gate circuits is configured by configuring the configuration file, thereby realizing the functions of some or all of the above units. All units of the above communication device can be implemented in the form of software called by the processor, or in the form of hardware circuits, or in part by the form of software called by the processor, and the rest by hardware circuits.

[0278] In the embodiments described in Figures 8 and 9, the processor is a circuit with signal processing capabilities. In one implementation, the processor can be a circuit with instruction reading and execution capabilities, such as a central processing unit (CPU), a microprocessor, a graphics processing unit (GPU) (which can be understood as a microprocessor), or a digital signal processor (DSP). In another implementation, the processor can implement certain functions through the logical relationship of a hardware circuit, and the logical relationship of the hardware circuit is fixed or reconfigurable. For example, the processor is a hardware circuit implemented by an application-specific integrated circuit (ASIC) or a programmable logic device (PLD), such as an FPGA. In a reconfigurable hardware circuit, the process of the processor loading a configuration document to implement the hardware circuit configuration can be understood as the process of the processor loading instructions to implement the functions of some or all of the above units. In addition, it can also be a hardware circuit designed for artificial intelligence, which can be understood as an ASIC, such as a neural network processing unit (NPU), a tensor processing unit (TPU), a deep learning processing unit (DPU), etc.

[0279] It can be seen that each unit in the above communication device can be one or more processors (or processing circuits) configured to implement the above method, such as: CPU, GPU, NPU, TPU, DPU, microprocessor, DSP, ASIC, FPGA, or a combination of at least two of these processor forms.

[0280] In addition, the various units in the above communication device can be fully or partially integrated together, or can be implemented independently. In one implementation, these units are integrated together and implemented in the form of a system-on-a-chip (SOC). The SOC may include at least one processor for implementing any of the above methods or implementing the functions of each unit of the communication device. The type of the at least one processor can be different, for example, including a CPU and FPGA, a CPU and an artificial intelligence processor, a CPU and a GPU, etc.

[0281] Optionally, the processor 901 may be a processor specifically used to execute these methods (referred to as a dedicated processor for ease of distinction), or may be a processor that executes these methods by calling a computer program, such as a general-purpose processor. Optionally, the at least one processor may include both a dedicated processor and a general-purpose processor. Optionally, when the communication device 90 includes at least one processor 901, the computer program may be stored in the memory 903.

[0282] The communication interface 902 can be used to provide information input or output for the at least one processor. And / or, the communication interface 902 can also be used to receive data sent externally and / or send data externally. It can be a wired link interface such as an Ethernet cable, or a wireless link interface (wireless fidelity (WiFi), Bluetooth, general wireless transmission, vehicle-mounted short-range communication technology, and other short-range wireless communication technologies). Optionally, the communication interface 902 can also include a transmitter (such as a radio frequency transmitter, antenna, etc.) coupled to the interface, or a receiver.

[0283] The memory 903 is used to provide storage space, in which data such as an operating system and computer programs can be stored. The memory 803 can be a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), or a portable read-only memory (CD-ROM), etc., or a combination of multiple types. The at least one processor 901 in the communication device 90 is used to execute the aforementioned communication method, such as the steps performed by the network device in the embodiment described in Figure 2. The relevant content can be referred to above and will not be repeated here.

[0284] The present application also provides a terminal, which includes the communication device as shown in Figure 6 or Figure 8.

[0285] The present application also provides a network device, which includes the communication device as shown in Figure 7 or Figure 9.

[0286] The present application also provides a communication system, which includes a terminal and a network device, wherein: the terminal includes the communication device as described in Figure 6 or Figure 8; the network device includes the communication device as described in Figure 7 or Figure 9.

[0287] The present application also provides a computer-readable storage medium, in which instructions are stored. When the instructions are executed on at least one processor, the aforementioned communication method, such as the method described in FIG. 2 , is implemented.

[0288] The present application also provides a computer program product, which includes computer instructions and, when executed by a computing device, implements the aforementioned communication method, such as the method described in FIG. 2 .

[0289] In the embodiments of this application, words such as "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described in this application as "exemplary" or "for example" should not be construed as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0290] The “at least one” mentioned in the embodiments of this application refers to one or more, and “plurality” refers to two or more. “At least one of the following items” or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can represent: a, b, c, (a and b), (a and c), (b and c), or (a and b and c), where a, b, c can be single or multiple. “And / or” describes the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The character “ / ” generally indicates that the previous and next associated objects are in an “or” relationship.

[0291] Furthermore, unless otherwise specified, ordinal numbers such as "first" and "second" in the embodiments of this application are used to distinguish multiple objects and are not used to limit the order, timing, priority, or importance of multiple objects. For example, the first device and the second device are only for ease of description and do not indicate differences in structure, importance, etc. between the first and second devices. In some embodiments, the first device and the second device can also be the same device.

[0292] In the above embodiments, the term "when" can be interpreted to mean "if...", "after...", "in response to determining...", or "in response to detecting...", depending on the context. The above are merely optional embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the concepts and principles of the present application shall be included in the scope of protection of the present application.

[0293] Those skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware, or by a program to instruct the relevant hardware, and the program may be stored in a computer-readable storage medium, which may be a read-only memory, a disk, or an optical disk, etc.

[0294] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A communication method, characterized in that: The method comprises: Acquire learning-related status information associated with an artificial intelligence (AI) task, wherein the learning-related status information includes one or more of resource status information, environment status information, model performance information, and data set status information; The learning-related status information is sent, wherein the learning-related status information is used to perform lifecycle management on the model associated with the AI ​​task.

2. The method according to claim 1, characterized in that Before acquiring the learning-related status information, the method further includes: Receive indication information, wherein the indication information is used to indicate a sending method of the learning-related status information.

3. The method according to claim 2, characterized in that The indication information includes a reference time; The sending of the learning-related status information includes: When the reference time is met, the learning-related status information is sent.

4. The method according to claim 2, characterized in that The indication information includes a trigger condition; the sending of the learning-related status information includes: When the trigger condition is met, the learning-related status information is sent.

5. The method according to any one of claims 1 to 4, characterized in that: The method further comprises: Sending indication information of a task identifier, wherein the indication information of the task identifier is used to indicate the associated AI task.

6. The method according to any one of claims 1 to 5, characterized in that: The resource status information includes resource margin, and the resource margin includes computing power margin and / or storage margin.

7. The method according to claim 6, characterized in that The resource status information includes multiple task identifiers and multiple resource margins, and each of the multiple task identifiers is used to represent an AI task associated with a resource margin among the multiple resource margins.

8. The method according to claim 7, characterized in that The resource status information is carried in the signaling of the media access control-control element MAC-CE, and the signaling of the MAC-CE includes a task identification field and a resource margin CPH field, the task identification field includes the task identification, and the CPH field includes the resource margin.

9. The method according to claim 8, characterized in that The method further comprises: Sending first indication information, wherein the first indication information is used to indicate that the learning-related status information comes from a first category of AI tasks or a second category of AI tasks, the first category of AI tasks includes one or more preset tasks, and the second category of AI tasks includes one or more AI tasks other than the first category of AI tasks.

10. The method according to claim 9, characterized in that The first indication information is a task identifier.

11. The method according to any one of claims 1 to 10, characterized in that: The environmental status information includes first parameter information, and the first parameter information includes one or more of a system parameter, a scene type, and a model identifier.

12. The method according to claim 11, characterized in that The environmental status information includes multiple task identifiers and multiple first parameter information, and each of the multiple task identifiers is used to represent an AI task associated with one of the multiple first parameter information.

13. The method according to claim 12, characterized in that The environmental status information is carried in the signaling of MAC-CE, and the signaling of MAC-CE includes a task identification field, a system parameter field, a scenario type field and a model identification field, the task identification field includes the task identification, the system parameter field includes the system parameter, the scenario type field includes the scenario type, and the model identification field includes the model identification.

14. The method according to any one of claims 1 to 13, characterized in that: The model performance information includes second parameter information, and the second parameter information includes one or more of a model identifier, a model performance, a fallback performance, and a fallback indication, wherein the model performance is used to characterize the performance of the model used for the associated AI task, the fallback performance is used to characterize the system performance of the associated AI task after the model execution falls back to the non-model execution, the fallback indication is used to characterize whether the associated AI task needs to be executed back, and the model identifier includes the model identifier of one or more models in the model associated with the execution of the AI ​​task.

15. The method according to claim 14, characterized in that The model performance information includes multiple task identifiers and multiple second parameter information, and each task identifier in the multiple task identifiers is used to represent an AI task associated with a second parameter information in the multiple second parameter information.

16. The method according to claim 15, characterized in that The model performance information is carried in the signaling of MAC-CE, and the signaling of MAC-CE includes a task identification field, a model identification field, a model performance field, a fallback performance field and a fallback indication field, the task identification field includes the task identification, the model identification field includes the model identification, the model performance field includes the model performance, the fallback performance field includes the fallback performance, and the fallback indication field includes the fallback indication.

17. The method according to any one of claims 1 to 16, characterized in that: The data set status information includes third parameter information, and the third parameter information includes one or more of data set size, data collection time, data collection location and data distribution description, wherein the data collection time includes the collection time of the latest data sample in the data set collected for the associated AI task and the collection time of the oldest data sample in the data set, the data set size is used to characterize the number of data samples in the data set collected for the associated AI task, the data collection location is used to characterize the location of data in the data set collected for the associated AI task, and the data distribution description is used to characterize the distribution of data in the data set collected for the associated AI task.

18. The method according to claim 17, characterized in that The data set status information includes multiple task identifiers and multiple third parameter information, and each of the multiple task identifiers is used to represent an AI task associated with one of the multiple third parameter information.

19. The method according to claim 18, characterized in that The data set status information is carried in the signaling of MAC-CE, and the signaling of MAC-CE includes a task identification field, a data set size field, a latest data collection time field, an oldest data collection time field, a data collection location field and a data distribution description field. The task identification field includes the task identification, the data set size field includes the data set size, the latest data collection time field includes the latest data collection time, the oldest data collection time field includes the oldest data collection time, the data collection location field includes the data collection location, and the data distribution description field includes the data distribution description.

20. A communication method, characterized in that: The method comprises: Receiving learning-related status information associated with an artificial intelligence (AI) task, wherein the learning-related status information includes one or more of resource status information, environment status information, model performance information, and data set status information; Perform lifecycle management on the model associated with the AI ​​task according to the learning-related status information.

21. The method according to claim 20, characterized in that Before receiving the learning-related status information, the method further includes: Sending indication information, wherein the indication information is used to indicate a sending method of the learning-related status information.

22. The method according to claim 21, characterized in that The indication information includes a reference time, and the indication information is used to indicate that the learning-related state information is sent when the reference time is met.

23. The method according to claim 21, characterized in that The indication information includes a trigger condition, and the indication information is used to indicate that the learning-related state information is sent when the trigger condition is met.

24. The method according to any one of claims 20 to 23, characterized in that: The method further comprises: Receive indication information of a task identifier, wherein the indication information of the task identifier is used to indicate the associated AI task.

25. The method according to any one of claims 20 to 24, characterized in that: The resource status information includes resource margin, and the resource margin includes computing power margin and / or storage margin.

26. The method according to claim 25, characterized in that The resource status information includes multiple task identifiers and multiple resource margins, and each of the multiple task identifiers is used to represent an AI task associated with a resource margin among the multiple resource margins.

27. The method according to claim 26, characterized in that The resource status information is carried in the signaling of the media access control-control element MAC-CE, and the signaling of the MAC-CE includes a task identification field and a resource margin CPH field, the task identification field includes the task identification, and the CPH field includes the resource margin.

28. The method according to claim 27, characterized in that The method further comprises: Receive first indication information, wherein the first indication information is used to indicate that the learning-related state information comes from a first category of AI tasks or a second category of AI tasks, the first category of AI tasks includes one or more preset tasks, and the second category of AI tasks includes one or more AI tasks other than the first category of AI tasks.

29. The method according to claim 28, characterized in that The first indication information is a task identifier.

30. The method according to any one of claims 20 to 29, characterized in that: The environmental status information includes first parameter information, and the first parameter information includes one or more of a system parameter, a scene type, and a model identifier.

31. The method according to claim 30, characterized in that The environmental status information includes multiple task identifiers and multiple first parameter information, and each of the multiple task identifiers is used to represent an AI task associated with one of the multiple first parameter information.

32. The method according to claim 31, characterized in that The environmental status information is carried in the signaling of MAC-CE, and the signaling of MAC-CE includes a task identification field, a system parameter field, a scenario type field and a model identification field, the task identification field includes the task identification, the system parameter field includes the system parameter, the scenario type field includes the scenario type, and the model identification field includes the model identification.

33. The method according to any one of claims 20 to 32, characterized in that: The model performance information includes second parameter information, and the second parameter information includes one or more of a model identifier, a model performance, a fallback performance, and a fallback indication, wherein the model performance is used to characterize the performance of the model used for the associated AI task, the fallback performance is used to characterize the system performance of the associated AI task after the model execution falls back to the non-model execution, the fallback indication is used to characterize whether the associated AI task needs to be executed back, and the model identifier includes the model identifier of one or more models in the model associated with the execution of the AI ​​task.

34. The method according to claim 33, characterized in that The model performance information includes multiple task identifiers and multiple second parameter information, and each task identifier in the multiple task identifiers is used to represent an AI task associated with a second parameter information in the multiple second parameter information.

35. The method according to claim 34, characterized in that The model performance information is carried in the signaling of MAC-CE, and the signaling of MAC-CE includes a task identification field, a model identification field, a model performance field, a fallback performance field and a fallback indication field, the task identification field includes the task identification, the model identification field includes the model identification, the model performance field includes the model performance, the fallback performance field includes the fallback performance, and the fallback indication field includes the fallback indication.

36. The method according to any one of claims 20 to 35, characterized in that The data set status information includes third parameter information, and the third parameter information includes one or more of data set size, data collection time, data collection location and data distribution description, wherein the data collection time includes the collection time of the latest data sample in the data set collected for the associated AI task and the collection time of the oldest data sample in the data set, the data set size is used to characterize the number of data samples in the data set collected for the associated AI task, the data collection location is used to characterize the location of data in the data set collected for the associated AI task, and the data distribution description is used to characterize the distribution of data in the data set collected for the associated AI task.

37. The method according to claim 36, characterized in that The data set status information includes multiple task identifiers and multiple third parameter information, and each of the multiple task identifiers is used to represent an AI task associated with one of the multiple third parameter information.

38. The method according to claim 37, characterized in that The data set status information is carried in the signaling of MAC-CE, and the signaling of MAC-CE includes a task identification field, a data set size field, a latest data collection time field, an oldest data collection time field, a data collection location field and a data distribution description field. The task identification field includes the task identification, the data set size field includes the data set size, the latest data collection time field includes the latest data collection time, the oldest data collection time field includes the oldest data collection time, the data collection location field includes the data collection location, and the data distribution description field includes the data distribution description.

39. A communication device, characterized in that: The communication device comprises a module for implementing the method according to any one of claims 1 to 19.

40. A communication device, characterized in that: The communication device comprises a module for implementing the method according to any one of claims 20 to 38.

41. A communication device, characterized in that: The communication device includes a processor; When the processor calls the computer program or instruction in the memory, the method according to any one of claims 1 to 19 is implemented.

42. A communication device, characterized in that: The communication device includes a processor; When the processor calls the computer program or instruction in the memory, the method according to any one of claims 20 to 38 is implemented.

43. A communication device, characterized in that: comprising a logic circuit and an interface, wherein the logic circuit and the interface are coupled; The interface is used to input and / or output information, and the logic circuit is used to execute the method according to any one of claims 1-38.

44. The device according to any one of claims 41 to 43, characterized in that The communication device is a chip or a chip system.

45. A communication system, characterized in that: The communication system comprises the communication device according to claim 39 and the communication device according to claim 40; or The communication system includes the communication device according to claim 41 and the communication device according to claim 42.

46. ​​A computer-readable storage medium, characterized in that The computer readable storage medium is used to store instructions or computer programs; When the instructions or the computer program are executed, the method according to any one of claims 1 to 38 is implemented.

47. A computer program product, characterized in that include: instructions or computer programs; When the instructions or the computer program are executed, the method according to any one of claims 1 to 38 is implemented.

Images