Wireless communication methods, devices, and storage medium

By instructing the AI ​​unit's storage status through the terminal device, the problem of unreasonable resource management when deploying AI models on the terminal device is solved, achieving efficient resource utilization and reasonable timing design of AI models.

WO2026232345A1PCT designated stage Publication Date: 2026-11-12GUANGDONG OPPO MOBILE TELECOMMUNICATIONS CORP LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/CN2025/093242
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Filing Date
2025-05-07
Publication Date
2026-11-12

AI Technical Summary

Technical Problem

In existing technologies, terminal devices fail to effectively consider storage and deployment latency when deploying AI models, resulting in unreasonable resource management and potential waste.

Method used

The terminal device indicates the storage status of the AI ​​unit to the network device, enabling the network device to reserve appropriate latency so that the communication system can perform timing design and resource management for different AI models.

Benefits of technology

By indicating the storage status of AI units, network devices can make reasonable use of resources, avoid resource waste, and achieve effective AI model management and timing design.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN2025093242_12112026_PF_FP_ABST
    Figure CN2025093242_12112026_PF_FP_ABST
Patent Text Reader

Abstract

Wireless communication methods, devices, and a storage medium, relating to the technical field of communications. A method is executed by a terminal device. The method comprises: sending first information, the first information being used for indicating a storage state of one or more AI units deployed by the terminal device (210). The terminal device indicates the storage state of the AI unit to a network device, such that the network device can reserve an appropriate delay on the basis of the storage state of the AI unit, thereby facilitating timing design and management of different AI models by a communication system, rationally utilizing resources, and avoiding waste of resources.
Need to check novelty before this filing date? Find Prior Art

Description

Wireless communication methods, devices and storage media Technical Field

[0001] This application relates to the field of communication technology, and in particular to a wireless communication method, device and storage medium. Background Technology

[0002] Drawing on the rapid development of AI (Artificial Intelligence) technology, the integration of artificial intelligence with wireless communication technology is also entering the public eye. Examples include AI-based CSI (Channel State Information) feedback, beam management, and positioning technologies.

[0003] Related technologies, considering the computing power of UE (User Equipment), provide a reference for the deployment of AI functions and models on terminal devices. However, the UE's requirements for AI model storage and deployment latency still need further discussion and research. Summary of the Invention

[0004] This application provides a wireless communication method, device, and storage medium. The technical solution is as follows:

[0005] According to one aspect of the embodiments of this application, a wireless communication method is provided, the method being executed by a terminal device, the method comprising:

[0006] Send first information, which is used to indicate the storage status of one or more AI units deployed on the terminal device.

[0007] According to one aspect of the embodiments of this application, a wireless communication method is provided, the method being performed by a network device, the method comprising:

[0008] Receive first information, which is used to indicate the storage status of one or more AI units deployed on the terminal device.

[0009] According to one aspect of the embodiments of this application, a terminal device is provided, the terminal device comprising:

[0010] The sending module is used to send first information, which is used to indicate the storage status of one or more AI units deployed in the terminal device.

[0011] According to one aspect of the embodiments of this application, a network device is provided, the network device comprising:

[0012] A receiving module is configured to receive first information, which indicates the storage status of one or more AI units deployed on a terminal device.

[0013] According to one aspect of the embodiments of this application, a communication device is provided, the communication device including a processor and a memory, the memory storing a computer program, the processor executing the computer program to implement the above-described wireless communication method. The communication device is a terminal device, or the communication device is a network device.

[0014] According to one aspect of the embodiments of this application, a computer-readable storage medium is provided, wherein the storage medium stores a computer program for execution by a processor to implement the above-described wireless communication method.

[0015] According to one aspect of the embodiments of this application, a chip is provided, the chip including programmable logic circuits and / or program instructions, which, when the chip is running, are used to implement the above-described wireless communication method.

[0016] According to one aspect of the embodiments of this application, a computer program product is provided, the computer program product including computer instructions stored in a computer-readable storage medium, and a processor reading from the computer-readable storage medium and executing the computer instructions to implement the above-described wireless communication method.

[0017] The technical solutions provided in this application embodiment may have the following beneficial effects:

[0018] The terminal device indicates the storage status of the AI ​​unit to the network device, so that the network device can reserve appropriate latency based on the storage status of the AI ​​unit. This facilitates the timing design and management of different AI models by the communication system, makes reasonable use of resources, and avoids resource waste. Attached Figure Description

[0019] Figure 1 is a schematic diagram of a network architecture provided in one embodiment of this application;

[0020] Figure 2 is a flowchart of a wireless communication method provided in an embodiment of this application;

[0021] Figure 3 is a flowchart of a wireless communication method provided in another embodiment of this application;

[0022] Figure 4 is a flowchart of a wireless communication method provided in another embodiment of this application;

[0023] Figure 5 is a block diagram of a terminal device provided in an embodiment of this application;

[0024] Figure 6 is a block diagram of a network device provided in an embodiment of this application;

[0025] Figure 7 is a schematic diagram of the structure of a communication device provided in one embodiment of this application. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0027] The network architecture and business scenarios described in the embodiments of this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided in the embodiments of this application. As those skilled in the art will know, with the evolution of network architecture and the emergence of new business scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0028] Please refer to Figure 1, which shows a schematic diagram of a network architecture 100 provided in one embodiment of this application. The network architecture 100 may include: a terminal device 10, an access network device 20, and a core network element 30.

[0029] Terminal device 10 can refer to UE (User Equipment), STA (Station), access terminal, user unit, user station, mobile station, mobile station, remote station, remote terminal, mobile device, wireless communication device, user agent, or user equipment. In some embodiments, terminal device 10 can also be a cellular phone, cordless phone, SIP (Session Initiation Protocol) phone, WLL (Wireless Local Loop) station, PDA (Personal Digital Assistant), handheld device with wireless communication capabilities, computing device or other processing device connected to a wireless modem, vehicle-mounted device, wearable device, terminal device in 5GS (5th Generation System), or terminal device in the future evolved PLMN (Public Land Mobile Network), etc., and this application embodiment is not limited to these. For ease of description, the devices mentioned above are collectively referred to as terminal devices. The number of terminal devices 10 is usually multiple, and one or more terminal devices 10 can be distributed within the cell managed by each access network device 20. Terminal equipment can also be simply referred to as terminal or UE, the meaning of which can be understood by those skilled in the art.

[0030] Access network device 20 is a device deployed in an access network to provide wireless communication functionality to terminal device 10. Access network device 20 may include various forms of macro base stations, micro base stations, relay stations, APs (Access Points), etc. In systems employing different wireless access technologies, the name of the device with access network device functionality may differ; for example, in a 5G NR (New Radio) system, it is called gNodeB or gNB (Next Generation Node B). As communication technologies evolve, the name "access network device" may change. For ease of description, in this embodiment, the aforementioned devices providing wireless communication functionality to terminal device 10 are collectively referred to as access network devices. In some embodiments, a communication relationship can be established between terminal device 10 and core network element 30 through access network device 20. For example, in an LTE (Long Term Evolution) system, access network device 20 can be one or more eNodeBs within an EUTRAN (Evolved Universal Terrestrial Radio Access Network); in a 5G NR system, access network device 20 can be one or more gNBs within a RAN (Radio Access Network). In the embodiments of this application, unless otherwise specified, "network device" refers to access network device 20, such as a base station.

[0031] Core network element 30 is a network element deployed in the core network. Its main functions are to provide user connectivity, manage users, and bear services, serving as an interface to external networks. For example, core network elements in a 5G NR system may include AMF (Access and Mobility Management Function) entities, UPF (User Plane Function) entities, and SMF (Session Management Function) entities.

[0032] In some embodiments, the access network device 20 and the core network element 30 communicate with each other via some air interface technology, such as the NG interface in a 5G NR system. The access network device 20 and the terminal device 10 communicate with each other via some air interface technology, such as the Uu interface.

[0033] The "5G NR system" in this application embodiment can also be referred to as a 5G system or an NR system, but those skilled in the art will understand its meaning. The technical solutions described in this application embodiment can be applied to LTE systems, 5G NR systems, and subsequent evolution systems of 5G NR systems (such as B5G (Beyound 5G) systems, 6G systems (6th Generation System), and other communication systems such as NB-IoT (Narrow Band Internet of Things) systems. This application does not limit these applications.

[0034] In this embodiment, the network device can provide services to a cell. The terminal device communicates with the network device through the transmission resources (e.g., frequency domain resources, or spectrum resources) on the carrier used by the cell. The cell can be the cell corresponding to the network device (e.g., a base station). The cell can belong to a macro base station or to a base station corresponding to a small cell. The small cell can include: Metro Cell, Micro Cell, Pico Cell, Femto Cell, etc. These small cells have the characteristics of small coverage area and low transmission power, and are suitable for providing high-speed data transmission services.

[0035] Before introducing the technical solution of this application, some related technical knowledge involved in this application will be introduced and explained. The following related technologies are optional solutions and can be arbitrarily combined with the technical solutions of the embodiments of this application, all of which fall within the protection scope of the embodiments of this application. The embodiments of this application include at least some of the following contents.

[0036] 1. Artificial Intelligence

[0037] In recent years, artificial intelligence (AI) technology, relying on the development of different types of neural networks and deep learning algorithms, has been widely applied in various fields such as image, speech, and video processing. Typical neural network architectures include fully connected networks, convolutional neural networks (CNNs), recurrent neural networks (RNNs), and Transformer structures with self-attention mechanisms, which can accomplish different task objectives. Drawing on the rapid development of AI technology, the combination of artificial intelligence and wireless communication technology has also attracted widespread interest from academia and industry. Among related technologies, AI-based CSI feedback, beam management, and positioning technologies have all undergone extensive research, evaluation, and standardization work.

[0038] 2. CPU (CSI Processing Unit)

[0039] In related technologies, the terminal device will report the number N of CSI processing units (CPUs) that it supports and occupies simultaneously through UE capability. CPU The number of CSI calculations that a terminal device can perform simultaneously cannot exceed the number reported.

[0040] Similarly, the discussion in Release 19 introduced the concept of the AI ​​Processing Unit (APU), which measures the ability of a terminal device to simultaneously process AI / ML computations. This concept is still under discussion and no stable conclusions have been reached, but some preliminary design ideas have emerged. For example, the computational units / resources used for CSI processing in the APU are separate from the CPU processing units for non-AI tasks; different AI functions / tasks share the APU computational unit / resource pool, and when the number of AI functions / tasks reaches the maximum number of APU computational units / resources allowed by the terminal device, selection or abandonment will be based on the priority of the AI ​​functions / tasks. The design of the APU or similar concepts provides a UE (User Equipment) capability reference for deploying AI functions and models on terminal devices.

[0041] While the currently discussed APU concept provides a UE capability reference for deploying AI functions and models in terminal devices, it only considers the UE's computing power and does not take into account the UE's requirements for AI model storage and deployment latency. Given that different AI models have different structures and parameter counts, and may be stored in the UE's memory or used in GPU memory, the deployment latency of AI models may vary depending on the location of the model structure and parameters. Therefore, base stations and users need to estimate this latency and design appropriately to avoid potential timeline problems. This application proposes a model storage and latency management method to address this challenge.

[0042] Please refer to Figure 2, which shows a flowchart of a wireless communication method provided in an embodiment of this application. This method is performed by a terminal device. The method includes the following step 210.

[0043] Step 210: The terminal device sends first information, which is used to indicate the storage status of one or more AI units deployed on the terminal device.

[0044] In some embodiments, the first information is carried in the capability information of the terminal device; or, the first information is carried in the auxiliary information of the terminal device.

[0045] 1. AI Unit

[0046] In some embodiments, the AI ​​unit consists of an AI model or a set of parameters of an AI model deployed on a terminal device.

[0047] In one example, an AI unit corresponds to an AI model group deployed on a terminal device, where the AI ​​model group is a functional division based on the AI ​​models. In some embodiments, one AI unit corresponds to one or more AI model groups deployed on the terminal device. For example, a first AI unit corresponds to a first AI model group deployed on the terminal device, which is used to implement a first function. For instance, if the terminal device deploys three AI models, AI1, AI2, and AI3, where AI1 and AI2 are used to implement the first function and AI3 is used to implement a second function, then the first AI unit corresponds to the first AI model group, which includes AI1 and AI2. Alternatively, a first AI unit may correspond to a first AI model group and a second AI model group deployed on the terminal device, where the first AI model group is used to implement the first function and the second AI model group is used to implement the second function; in this case, the first AI unit includes AI1, AI2, and AI3.

[0048] In some embodiments, the functions implemented by the AI ​​model can also be referred to as the functional characteristics of the AI ​​model, or the tasks to which the AI ​​model is applicable. For ease of description, the embodiments of this application will be described using the functions implemented by the AI ​​model.

[0049] In another example, an AI unit corresponds to an AI model deployed on a terminal device. In some embodiments, one AI unit corresponds to one or more AI models deployed on a terminal device. For example, a first AI unit corresponds to a first AI model deployed on a terminal device. For instance, if the terminal device deploys three AI models, AI1, AI2, and AI3, the first AI unit corresponds to the first AI model, and the first AI model is any one or more of AI1, AI2, and AI3.

[0050] The AI ​​model mentioned in the embodiments of this application refers to the structure of an AI model. The same AI model structure may correspond to one parameter set or multiple parameter sets.

[0051] In another example, an AI unit corresponds to the parameter set of a first AI model, which is any AI model deployed on the terminal device. In some embodiments, an AI unit corresponds to one or more parameter sets of the first AI model. In some embodiments, for an AI model, the terminal device may store one or more parameter sets. For example, for AI1, the terminal device stores three parameter sets S1, S2, and S3, and the first AI unit corresponds to the first parameter set of the first AI model. For example, the first AI unit corresponds to parameter set S1, or, for another example, the first AI unit corresponds to parameter sets S1 and S2.

[0052] 2. AI Model

[0053] In some embodiments, an AI model is a mathematical model built based on artificial intelligence technology to simulate human thought processes or solve complex problems. Exemplarily, an AI model may include an ML model and a deep learning model. Of course, this application does not limit the structure of the AI ​​model; here, the model deployed on the terminal device to achieve a specific function is collectively referred to as an AI model. For example, an AI model is a model deployed on the terminal device for compressing or predicting CSI.

[0054] In some embodiments, the AI ​​model deployed on the terminal device includes: an AI model deployed on one side of the terminal device, and a portion of the AI ​​model deployed on both sides stored on the terminal device side.

[0055] 3. Storage Status

[0056] In some embodiments, the storage state can be divided into a first state and a second state depending on the storage location of the AI ​​unit.

[0057] In some embodiments, the AI ​​unit in the first state is stored in the computing unit of the terminal device or a storage unit associated with the computing unit. Since the AI ​​unit is stored in the computing unit of the terminal device or a storage unit associated with the computing unit, its model structure and parameters have already been compiled. When the AI ​​unit needs to perform inference, it can immediately enter the inference process. At this point, the AI ​​unit can be considered ready, or in other words, in an active state.

[0058] In some embodiments, the AI ​​unit in the second state is not stored in the computing unit or storage unit associated with the computing unit of the terminal device. Exemplarily, the AI ​​unit in the second state is stored in the memory of the terminal device, or in a server on the terminal device or chip side. In this state, when the AI ​​unit needs to perform inference, additional parameter loading and model compilation processes are required, thus requiring a certain delay before the inference process can begin. At this time, the AI ​​unit can be considered not ready, or in an inactive state.

[0059] It should be noted that in this embodiment, the description refers to the storage state of the AI ​​unit, which can be understood as the storage state of the AI ​​models included in the AI ​​unit. For example, if the first AI unit is in the first state, it means that all the AI ​​models included in the first AI unit are in the first state.

[0060] The technical solution provided in this application embodiment indicates the storage status of the AI ​​unit to the network device, so that the network device can reserve an appropriate delay according to the storage status of the AI ​​unit, which facilitates the timing design and management of different AI models by the communication system, makes reasonable use of resources, and avoids waste of resources.

[0061] Please refer to Figure 3, which shows a flowchart of a wireless communication method provided in an embodiment of this application. This method is performed by a network device. The method includes the following step 310.

[0062] Step 310: The network device receives first information, which is used to indicate the storage status of one or more AI units deployed on the terminal device.

[0063] In some embodiments, the first information is carried in the capability information of the terminal device; or, the first information is carried in the auxiliary information of the terminal device.

[0064] 1. AI Unit

[0065] In some embodiments, the AI ​​unit consists of an AI model or a set of parameters of an AI model deployed on a terminal device.

[0066] In one example, an AI unit corresponds to an AI model group deployed on a terminal device, where the AI ​​model group is a functional division based on the AI ​​models. In some embodiments, one AI unit corresponds to one or more AI model groups deployed on the terminal device. For example, a first AI unit corresponds to a first AI model group deployed on the terminal device, which is used to implement a first function. For instance, if the terminal device deploys three AI models, AI1, AI2, and AI3, where AI1 and AI2 are used to implement the first function and AI3 is used to implement a second function, then the first AI unit corresponds to the first AI model group, which includes AI1 and AI2. Alternatively, a first AI unit may correspond to a first AI model group and a second AI model group deployed on the terminal device, where the first AI model group is used to implement the first function and the second AI model group is used to implement the second function; in this case, the first AI unit includes AI1, AI2, and AI3.

[0067] In some embodiments, the functions implemented by the AI ​​model can also be referred to as the functional characteristics of the AI ​​model, or the tasks to which the AI ​​model is applicable. For ease of description, the embodiments of this application will be described using the functions implemented by the AI ​​model.

[0068] In another example, an AI unit corresponds to an AI model deployed on a terminal device. Exemplarily, a first AI unit corresponds to a first AI model deployed on the terminal device. In some embodiments, one AI unit corresponds to one or more AI models deployed on the terminal device. For example, if the terminal device deploys three AI models, AI1, AI2, and AI3, the first AI unit corresponds to the first AI model, which is any one or more of AI1, AI2, and AI3.

[0069] The AI ​​model mentioned in the embodiments of this application refers to the structure of an AI model. The same AI model structure may correspond to one parameter set or multiple parameter sets.

[0070] In another example, an AI unit corresponds to the parameter set of a first AI model, which is any AI model deployed on the terminal device. In some embodiments, an AI unit corresponds to one or more parameter sets of the first AI model. In some embodiments, for an AI model, the terminal device may store one or more parameter sets. For example, for AI1, the terminal device stores three parameter sets S1, S2, and S3, and the first AI unit corresponds to the first parameter set of the first AI model. For example, the first AI unit corresponds to parameter set S1, or, for another example, the first AI unit corresponds to parameter sets S1 and S2.

[0071] 2. AI Model

[0072] In some embodiments, an AI model is a mathematical model built based on artificial intelligence technology to simulate human thought processes or solve complex problems. Exemplarily, an AI model may include an ML model and a deep learning model. Of course, this application does not limit the structure of the AI ​​model; here, the model deployed on the terminal device to achieve a specific function is collectively referred to as an AI model. For example, an AI model is a model deployed on the terminal device for compressing or predicting CSI.

[0073] In some embodiments, the AI ​​model deployed on the terminal device includes: an AI model deployed on one side of the terminal device, and a portion of the AI ​​model deployed on both sides stored on the terminal device side.

[0074] 3. Storage Status

[0075] In some embodiments, the storage state can be divided into a first state and a second state depending on the storage location of the AI ​​unit.

[0076] In some embodiments, the AI ​​unit in the first state is stored in the computing unit of the terminal device or a storage unit associated with the computing unit. Since the AI ​​unit is stored in the computing unit of the terminal device or a storage unit associated with the computing unit, its model structure and parameters have already been compiled. When the AI ​​unit needs to perform inference, it can immediately enter the inference process. At this point, the AI ​​unit can be considered ready, or in other words, in an active state.

[0077] In some embodiments, the AI ​​unit in the second state is not stored in the computing unit or storage unit associated with the computing unit of the terminal device. Exemplarily, the AI ​​unit in the second state is stored in the memory of the terminal device, or in a server on the terminal device or chip side. In this state, when the AI ​​unit needs to perform inference, additional parameter loading and model compilation processes are required, thus requiring a certain delay before the inference process can begin. At this time, the AI ​​unit can be considered not ready, or in an inactive state.

[0078] It should be noted that in this embodiment, the description refers to the storage state of the AI ​​unit, which can be understood as the storage state of the AI ​​models included in the AI ​​unit. For example, if the first AI unit is in the first state, it means that all the AI ​​models included in the first AI unit are in the first state.

[0079] The technical solution provided in this application embodiment indicates the storage status of the AI ​​unit to the network device, so that the network device can reserve an appropriate delay according to the storage status of the AI ​​unit, which facilitates the timing design and management of different AI models by the communication system, makes reasonable use of resources, and avoids waste of resources.

[0080] In some embodiments, a number of AI models may be deployed in the terminal device. For example, the terminal device may deploy AI models to implement P1 functions, each function corresponds to P2 AI models, and each AI model corresponds to P3 parameter sets, where P1, P2, and P3 are positive integers. When a number of AI models are deployed in the terminal device, how to reduce the amount of data in the first information also needs to be considered. The above embodiments show that the first information indicates the storage state of the AI ​​model based on AI units. Specifically, how the first information indicates the storage state of the AI ​​unit will be explained in the following embodiments.

[0081] 1. The structure of the first piece of information

[0082] In some embodiments, the first information indicates the storage state of one or more AI units via one or more bits.

[0083] In one example, the first piece of information indicates the storage state of one or more AI units via a single bit.

[0084] For example, the first information indicates the storage state of the first AI unit through a bit. If the value of the bit is 1, it indicates that the first AI unit is in a first state; if the value of the bit is 0, it indicates that the first AI unit is in a second state. Alternatively, if the value of the bit is 0, it indicates that the first AI unit is in a first state; if the value of the bit is 1, it indicates that the first AI unit is in a second state.

[0085] For example, the first information indicates the storage state of the first AI unit group through a single bit. If the value of this bit is 1, it indicates that the first AI unit group is in a first state; if the value of this bit is 0, it indicates that the first AI unit group is in a second state. Alternatively, if the value of this bit is 0, it indicates that the first AI unit group is in a first state; if the value of this bit is 1, it indicates that the first AI unit group is in a second state. Here, the first AI unit group includes multiple AI units. The first AI unit group being in a first state means that all the AI ​​units included in the first AI unit group are in a first state; the first AI unit group being in a second state means that all the AI ​​units included in the first AI unit group are in a second state.

[0086] In another example, the first information indicates the storage state of multiple AI units through a bitmap, where one bit in the bitmap corresponds to one AI unit among the multiple AI units.

[0087] In some embodiments, where the first information is used to indicate the storage state of multiple AI units, the storage states of the multiple AI units may be the same or different.

[0088] For example, the first information indicates the storage state of multiple AI units through a bitmap. The bitmap includes 5 bits, which correspond to AI units 1 to 5 respectively. For instance, AI units 1 to 3 are in the first state, and AI units 4 and 5 are in the second state. If a bit value of 1 indicates that the corresponding AI unit is in the first state, and a bit value of 0 indicates that the corresponding AI unit is in the second state, then the bitmap is 11100; if a bit value of 0 indicates that the corresponding AI unit is in the first state, and a bit value of 1 indicates that the corresponding AI unit is in the second state, then the bitmap is 00011.

[0089] 2. Division of AI Units

[0090] The above embodiments describe how AI units can be divided based on the functions implemented by the AI ​​models, the structure of the AI ​​models, and the parameter sets of the AI ​​models. However, in reality, different AI models for the same function may all be in the first state, or all in the second state, or some AI models may be in the first state while others are in the second state. Furthermore, an AI model comprises two parts: the first part is the model structure, and the second part is the parameter set. The same model structure may correspond to one parameter set or multiple parameter sets. For multiple parameter sets under the same AI model, they may all be in the first state, or all in the second state, or some parameter sets may be in the first state while others are in the second state.

[0091] For example, if the AI ​​units are divided according to the functions implemented by the AI ​​models, and the AI ​​models in the first AI unit are used to implement the first function, then the first information indicates that all AI models and all parameter sets in the first AI unit are in the corresponding state.

[0092] For example, if the AI ​​units are divided according to the structure of the AI ​​models, and the first AI unit is a certain AI model or a group of AI models, then the first information indicates that all AI models and all parameter sets in the first AI unit are in the corresponding state.

[0093] It is feasible. If the AI ​​units are divided according to parameter sets, and the first AI unit is a certain parameter set or a group of parameter sets, then only the parameter set indicated by the first information is in the corresponding state.

[0094] 3. Storage of AI models

[0095] Considering the capabilities of the computing unit or the storage unit associated with the computing unit of the terminal device, and the fact that the storage space occupied by the AI ​​model in its first state cannot be too large, otherwise it may affect other functions of the terminal device.

[0096] In some embodiments, the storage space occupied by the AI ​​unit in the first state is less than or equal to a first threshold.

[0097] In some embodiments, in the first state, the storage space occupied by the AI ​​unit used to implement the first function is less than or equal to a second threshold.

[0098] For example, the storage space occupied by the AI ​​unit in the first state is less than or equal to Y1; and the storage space occupied by the AI ​​unit used to implement the first function in the first state is less than or equal to Y2. Wherein, Y1 and Y2 are positive integers.

[0099] In some embodiments, different AI models may occupy the same or different amounts of storage space. For example, different AI models may have different numbers of parameters, and therefore occupy different amounts of storage space. When an AI model is associated with multiple parameter sets, it also requires more storage space.

[0100] In some embodiments, the first threshold is reported by the terminal device to the network device, or the first threshold and the second threshold are reported by the terminal device to the network device. Exemplarily, the first information is also used to indicate the first threshold and / or the second threshold. If the terminal device does not report the second threshold, there is no constraint on the storage space occupied by the AI ​​unit used to implement the first function.

[0101] In some embodiments, the first threshold is predefined or preconfigured, or both the first and second thresholds are predefined or preconfigured. If the second threshold is not configured, there is no constraint on the storage space occupied by the AI ​​unit used to implement the first function.

[0102] In some embodiments, the second thresholds corresponding to AI units that implement different functions may be the same or different. For example, the second threshold corresponding to the AI ​​unit that implements the first function is 5M, and the second threshold corresponding to the AI ​​unit that implements the second function is also 5M. For example, the second threshold corresponding to the AI ​​unit that implements the first function is 5M, and the second threshold corresponding to the AI ​​unit that implements the second function is 12M.

[0103] In some embodiments, the storage space occupied by the AI ​​unit can be indicated by the size of the storage space or by the number of storage units occupied by the AI ​​unit.

[0104] In one example, the storage space occupied by the first AI unit is determined by the size of the storage space occupied by the first AI unit. For example, the storage space size is explicitly indicated, that is, the storage space is equal to the space occupied by the first AI unit, for example, the two AI units occupy 5M and 12M of storage space respectively.

[0105] In another example, the storage space occupied by the first AI unit is determined by the number of storage units occupied by the first AI unit, where the size of one storage unit is predefined or preconfigured; the first AI unit is any one of one or more AI units. For example, the storage space size is implicitly indicated, i.e., the storage space occupied by the AI ​​unit is quantified, such as 5MB per storage unit, then the two AI units occupy 1 and 3 storage units respectively.

[0106] 4. Storage status update

[0107] In some embodiments, as shown in FIG4, the method further includes the following step 220.

[0108] In step 220, the terminal device sends second information, which is used to update the storage state of one or more AI units, or to indicate AI units whose storage state has changed. Correspondingly, the network device receives the second information.

[0109] When the first AI unit switches from a first state to a second state, the terminal device needs to report that the first AI unit is in the second state. When the first AI unit is in the second state, it needs to be activated and inference performed; therefore, the network device considers that the terminal device needs to activate the latency of the first AI unit. When the first AI unit switches from a second state to a first state, the terminal device needs to report that the first AI unit is in the first state. When the first AI unit is in the first state, it does not need to be activated and inference performed; therefore, the network device considers that the terminal device does not need to activate the latency of the first AI unit. Timely updating the storage state of the first AI unit allows for more efficient and rational use of resources.

[0110] In one example, the second information is used to update the storage state of one or more AI units. For example, when a first AI unit switches from a first state to a second state, the second information indicates that the storage state of the first AI unit has switched to the second state. Alternatively, when a first AI unit switches from a second state to a first state, the second information indicates that the storage state of the first AI unit has switched to the first state.

[0111] In another example, the second information is used to indicate the AI ​​unit whose storage state has changed. Exemplarily, the second information is used to indicate the identification information of the AI ​​unit whose storage state has changed. For example, the second information is used to indicate the index of the first AI unit. After receiving the second information, the network device determines that the storage state of the first AI unit has changed. In some embodiments, the network device maintains a state table corresponding to one or more AI units, the state table storing the storage states of the one or more AI units. After receiving the second information, the network device flips the storage state of the first AI unit stored in the state table. Flipping the storage state of the first AI unit stored in the state table means switching the first state to the second state, or switching the second state to the first state.

[0112] In some embodiments, if the first AI unit is in the second state, when the first AI unit is activated and inference is performed, the first AI unit will still fall back to the second state after inference, that is, the model structure and parameter set are cleared from the computing unit or the storage unit associated with the computing unit. At this time, the terminal device does not need to report the second information.

[0113] In some embodiments, if the first AI unit is in the second state, when the first AI unit is activated and inference is performed, the first AI unit maintains the first state after inference, that is, the model structure and parameter set are kept in the ready state in the computing unit or the storage unit associated with the computing unit. At this time, the terminal device needs to report the second information to indicate that the first AI unit is in the first state.

[0114] 5. Delay Information

[0115] In some embodiments, if the first AI unit is in the second state, the first information is further used to indicate delay information, the delay information being used to determine the delay unit corresponding to the first AI unit, and the first AI unit being any one of one or more AI units.

[0116] In some embodiments, the delay in the second state is referred to as a delay unit. The time unit of the delay unit can be a millisecond, a second, a time slot, a frame, or other time units.

[0117] In some embodiments, the latency units corresponding to different AI units may be the same or different.

[0118] In some embodiments, the delay unit is the time interval between the second time moment and the first time moment, and the first time moment is the moment when the first AI unit enters the first state.

[0119] In one example, the second moment is the moment when the terminal device receives the first signaling, which is used to activate the first AI unit. Exemplarily, the first signaling can be RRC (Radio Resource Control), MAC CE (Media Access Control-Control Element), or DCI (Downlink Control Information).

[0120] In another example, the second moment is the moment when the terminal device receives the first channel, which carries the first signaling for activating the AI ​​unit. Exemplarily, the first channel may be a PDSCH (Physical Downlink Shared Channel) or PDCCH (Physical Downlink Control Channel) carrying DCI.

[0121] In another example, the second moment is the moment when the terminal device receives the first reference signal. Exemplarily, the first reference signal can be CSI-RS (Channel State Information Reference Signal), DMRS (Demodulation Reference Signal), PTRS (Phase-Tracking Reference Signal), etc.

[0122] In another example, the second moment is the moment when the terminal device receives the first reference resource. Exemplarily, the first reference resource may not carry information and may only be used as a reference for the start time of the calculation delay unit.

[0123] The delay unit and the function used by the AI ​​model and / or the relationship between the AI ​​model can be implemented in the following different ways:

[0124] The delay unit is associated with the function that the AI ​​model is used to implement: for example, the first function is associated with the first delay unit X1, and the second function is associated with the second delay unit X2. X1 and X2 can be the same or different.

[0125] The delay unit is associated with the AI ​​model: For example, the first AI model under the first function is associated with the first delay unit X1, and the second AI model under the first function is associated with the second delay unit X2. X1 and X2 can be the same or different.

[0126] Since the number of parameters in different parameter sets corresponding to the same AI model is the same, and the only difference is in the parameters, the storage space occupied by different parameter sets is the same. Therefore, the delay units corresponding to different parameter sets are also the same, and there is no need to configure delay units separately for each parameter set.

[0127] In some embodiments, the first information may be reported to the delay unit according to the function implemented by the delay unit and / or the association between the delay unit and the AI ​​model and the AI ​​model.

[0128] In one example, if the first AI unit is applied to an AI model group that implements a first function, latency information is used to indicate the latency unit associated with the first function. For example, for the first function, the terminal device reports the latency unit X associated with the first function.

[0129] In another example, if the first AI unit corresponds to the first AI model, the latency information is used to indicate the latency coefficient corresponding to the first AI model. In some embodiments, the latency unit associated with the first AI model is the product of the latency coefficient corresponding to the first AI model and the storage space occupied by the first AI model. For example, the terminal device reports a latency coefficient c, indicating that the latency unit of the terminal device is cY when a certain AI unit occupies a storage unit Y.

[0130] This application provides a method for managing the storage status and latency of AI models, enabling network devices and terminal devices to measure and align the model storage capabilities on the terminal device side, as well as the status and deployment latency of AI models to be activated, facilitating the system's timing design and management of different functions and AI models.

[0131] It should be noted that, in the above method embodiments, the steps performed by the terminal device can be implemented separately as a wireless communication method on the terminal device side; the steps performed by the network device can be implemented separately as a wireless communication method on the network device side.

[0132] The following are embodiments of the device described in this application, which can be used to execute the embodiments of the method described in this application. For details not disclosed in the embodiments of the device described in this application, please refer to the embodiments of the method described in this application.

[0133] Please refer to Figure 5, which shows a block diagram of a terminal device provided in an embodiment of this application. This terminal device has the functionality to implement the wireless communication method example described above. This functionality can be implemented in hardware or by hardware executing corresponding software. As shown in Figure 5, the terminal device 500 may include a transmitting module 510.

[0134] The sending module 510 is used to send first information, which is used to indicate the storage status of one or more artificial intelligence (AI) units deployed in the terminal device.

[0135] In some embodiments, the first information indicates the storage state of the one or more AI units through a single bit; or, the first information indicates the storage state of multiple AI units through a bitmap, wherein one bit in the bitmap corresponds to one of the multiple AI units.

[0136] In some embodiments, the AI ​​unit corresponds to an AI model group deployed on the terminal device, the AI ​​model group being a functional division based on AI models; or, the AI ​​unit corresponds to an AI model deployed on the terminal device; or, the AI ​​unit corresponds to a parameter set of a first AI model, the first AI model being any AI model deployed on the terminal device.

[0137] In some embodiments, where the first information is used to indicate the storage status of multiple AI units, the storage status of the multiple AI units may be the same or different.

[0138] In some embodiments, the AI ​​unit in the first state is stored in the computing unit of the terminal device or a storage unit associated with the computing unit; or, the AI ​​unit in the second state is not stored in the computing unit of the terminal device or a storage unit associated with the computing unit.

[0139] In some embodiments, the storage space occupied by the AI ​​unit in the first state is less than or equal to a first threshold; and / or, in the first state, the storage space occupied by the AI ​​unit for implementing the first function is less than or equal to a second threshold.

[0140] In some embodiments, the first information is also used to indicate the first threshold and / or the second threshold.

[0141] In some embodiments, the storage space occupied by the first AI unit is determined by the size of the storage space occupied by the first AI unit; or, the storage space occupied by the first AI unit is determined by the number of storage units occupied by the first AI unit, wherein the size of one of the storage units is predefined or preconfigured; wherein the first AI unit is any one of the one or more AI units.

[0142] In some embodiments, the sending module 510 is further configured to send second information, the second information being used to update the storage state of the one or more AI units, or the second information being used to indicate AI units whose storage state has changed.

[0143] In some embodiments, if the first AI unit is in the second state, the first information is further used to indicate latency information, the latency information being used to determine the latency unit corresponding to the first AI unit, wherein the first AI unit is any one of the one or more AI units.

[0144] In some embodiments, if the first AI unit is applied to an AI model group that implements a first function, the latency information is used to indicate the latency unit associated with the first function; or, if the first AI unit corresponds to a first AI model, the latency information is used to indicate the latency coefficient corresponding to the first AI model.

[0145] In some embodiments, the latency unit associated with the first AI model is the product of the latency coefficient corresponding to the first AI model and the storage space occupied by the first AI model.

[0146] In some embodiments, the delay unit is the time interval between the second time point and the first time point, where the first time point is the moment when the first AI unit enters the first state.

[0147] In some embodiments, the second time is the time when the terminal device receives the first signaling, the first signaling being used to activate the first AI unit; or, the second time is the time when the terminal device receives the first channel, the first channel carrying the first signaling for activating the AI ​​unit; or, the second time is the time when the terminal device receives the first reference signal; or, the second time is the time when the terminal device receives the first reference resource.

[0148] In some embodiments, the first information is carried in the capability information of the terminal device; or, the first information is carried in the auxiliary information of the terminal device.

[0149] The technical solution provided in this application embodiment indicates the storage status of the AI ​​unit to the network device, so that the network device can reserve an appropriate delay according to the storage status of the AI ​​unit, which facilitates the timing design and management of different AI models by the communication system, makes reasonable use of resources, and avoids waste of resources.

[0150] Please refer to Figure 6, which shows a block diagram of a network device provided in one embodiment of this application. This network device has the functionality to implement the wireless communication method example described above. This functionality can be implemented in hardware or by hardware executing corresponding software. As shown in Figure 6, the network device 600 may include a receiving module 610.

[0151] The receiving module 610 is used to receive first information, which is used to indicate the storage status of one or more artificial intelligence (AI) units deployed in the terminal device.

[0152] In some embodiments, the first information indicates the storage state of the one or more AI units through a single bit; or, the first information indicates the storage state of multiple AI units through a bitmap, wherein one bit in the bitmap corresponds to one of the multiple AI units.

[0153] In some embodiments, the AI ​​unit corresponds to an AI model group deployed on the terminal device, the AI ​​model group being a functional division based on AI models; or, the AI ​​unit corresponds to an AI model deployed on the terminal device; or, the AI ​​unit corresponds to a parameter set of a first AI model, the first AI model being any AI model deployed on the terminal device.

[0154] In some embodiments, where the first information is used to indicate the storage status of multiple AI units, the storage status of the multiple AI units may be the same or different.

[0155] In some embodiments, the AI ​​unit in the first state is stored in the computing unit of the terminal device or a storage unit associated with the computing unit; or, the AI ​​unit in the second state is not stored in the computing unit of the terminal device or a storage unit associated with the computing unit.

[0156] In some embodiments, the storage space occupied by the AI ​​unit in the first state is less than or equal to a first threshold; and / or, in the first state, the storage space occupied by the AI ​​unit for implementing the first function is less than or equal to a second threshold.

[0157] In some embodiments, the first information is also used to indicate the first threshold and / or the second threshold.

[0158] In some embodiments, the storage space occupied by the first AI unit is determined by the size of the storage space occupied by the first AI unit; or, the storage space occupied by the first AI unit is determined by the number of storage units occupied by the first AI unit, wherein the size of one of the storage units is predefined or preconfigured; wherein the first AI unit is any one of the one or more AI units.

[0159] In some embodiments, the receiving module 610 is further configured to receive second information, the second information being used to update the storage state of the one or more AI units, or the second information being used to indicate an AI unit whose storage state has changed.

[0160] In some embodiments, if the first AI unit is in the second state, the first information is further used to indicate latency information, the latency information being used to determine the latency unit corresponding to the first AI unit, wherein the first AI unit is any one of the one or more AI units.

[0161] In some embodiments, if the first AI unit is applied to an AI model group that implements a first function, the latency information is used to indicate the latency unit associated with the first function; or, if the first AI unit corresponds to a first AI model, the latency information is used to indicate the latency coefficient corresponding to the first AI model.

[0162] In some embodiments, the latency unit associated with the first AI model is the product of the latency coefficient corresponding to the first AI model and the storage space occupied by the first AI model.

[0163] In some embodiments, the delay unit is the time interval between the second time point and the first time point, where the first time point is the moment when the first AI unit enters the first state.

[0164] In some embodiments, the second time is the time when the terminal device receives the first signaling, the first signaling being used to activate the first AI unit; or, the second time is the time when the terminal device receives the first channel, the first channel carrying the first signaling for activating the AI ​​unit; or, the second time is the time when the terminal device receives the first reference signal; or, the second time is the time when the terminal device receives the first reference resource.

[0165] In some embodiments, the first information is carried in the capability information of the terminal device; or, the first information is carried in the auxiliary information of the terminal device.

[0166] The technical solution provided in this application embodiment indicates the storage status of the AI ​​unit to the network device, so that the network device can reserve an appropriate delay according to the storage status of the AI ​​unit, which facilitates the timing design and management of different AI models by the communication system, makes reasonable use of resources, and avoids waste of resources.

[0167] It should be noted that the devices (terminal devices and network devices) provided in the above embodiments are only illustrated by the division of the above functional modules when implementing their functions. In actual applications, the above functions can be assigned to different functional modules according to actual needs, that is, the content structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0168] Regarding the device in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0169] Please refer to Figure 7, which shows a schematic diagram of a communication device provided in one embodiment of this application. The communication device can be a terminal device or a network device as described above. The communication device 700 may include a processor 701, a transceiver 702, and a memory 703. The transceiver 702 is used to implement sending or receiving functions, such as implementing the functions of the sending module 510 or the receiving module 610 described above. The processor 701 can be used to implement other processing functions or control sending and / or receiving.

[0170] The processor 701 includes one or more processing cores. The processor 701 executes various functional applications and information processing by running software programs and modules.

[0171] The transceiver 702 may include a receiver and a transmitter. For example, the receiver and transmitter may be implemented as the same wireless communication component, which may include a wireless communication chip and a radio frequency antenna.

[0172] The memory 703 can be connected to the processor 701 and the transceiver 702.

[0173] The memory 703 can be used to store a computer program executed by the processor, and the processor 701 is used to execute the computer program to implement the various steps in the above method embodiments.

[0174] In some embodiments, when the communication device 700 is a terminal device, the transceiver 702 is used to send first information, which is used to indicate the storage status of one or more artificial intelligence (AI) units deployed on the terminal device.

[0175] In some embodiments, when the communication device 700 is a network device, the transceiver 702 is used to receive first information, which is used to indicate the storage status of one or more artificial intelligence (AI) units deployed on the terminal device.

[0176] For details not described in this embodiment, please refer to the embodiments above, which will not be repeated here.

[0177] Furthermore, the memory can be implemented by any type of volatile or non-volatile storage device or a combination thereof, including but not limited to: magnetic disks or optical disks, electrically erasable programmable read-only memory, erasable programmable read-only memory, statically accessible memory, read-only memory, magnetic memory, flash memory, and programmable read-only memory.

[0178] This application embodiment also provides a computer-readable storage medium storing a computer program. The computer program is executed by a processor to implement the aforementioned wireless communication method on the terminal device side or the aforementioned wireless communication method on the network device side. Optionally, the computer-readable storage medium may include: ROM (Read-Only Memory), RAM (Random-Access Memory), SSD (Solid State Drives), or optical disc, etc. The random access memory may include ReRAM (Resistance Random Access Memory) and DRAM (Dynamic Random Access Memory).

[0179] This application also provides a chip, which includes programmable logic circuits and / or program instructions. When the chip is running, it is used to implement the wireless communication method on the terminal device side or the wireless communication method on the network device side.

[0180] This application also provides a computer program product, which includes a computer program stored in a computer-readable storage medium. A processor reads and executes the computer program from the computer-readable storage medium to implement the wireless communication method on the terminal device side or the wireless communication method on the network device side.

[0181] It should be understood that the term "instruction" mentioned in the embodiments of this application can be a direct instruction, an indirect instruction, or an indication of a relationship. For example, A instructing B can mean that A directly instructs B, such as B being able to obtain information through A; it can also mean that A indirectly instructs B, such as A instructing C, so B can obtain information through C; or it can mean that there is a relationship between A and B.

[0182] In the description of the embodiments of this application, the term "correspondence" may indicate that there is a direct or indirect correspondence between two things, or that there is an association between two things, or that there is a relationship of instruction and being instructed, configuration and being configured, etc.

[0183] In some embodiments of this application, "predefined" can be implemented by pre-storing corresponding codes, tables, or other means that can be used to indicate relevant information in the device (e.g., including terminal devices and APs). This application does not limit the specific implementation method. For example, predefined can refer to what is defined in the protocol.

[0184] In some embodiments of this application, the term "protocol" may refer to standard protocols in the field of communications, such as LTE protocols, NR protocols, and related protocols applied in future communication systems. This application does not limit the scope of these protocols.

[0185] In this article, "multiple" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0186] In this article, "greater than or equal to" can mean greater than or equal to, and "less than or equal to" can mean less than or equal to.

[0187] Furthermore, the step numbers described herein are merely illustrative of one possible execution order between steps. In some other embodiments, the steps may not be executed in the order of their numbers, such as two steps with different numbers being executed simultaneously, or two steps with different numbers being executed in the reverse order of the illustration. This application does not limit this.

[0188] Those skilled in the art will recognize that the functions described in the embodiments of this application in one or more of the above examples can be implemented using hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or code on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any medium that facilitates the transfer of a computer program from one place to another. Storage media can be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0189] The above description is merely an exemplary embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A wireless communication method, characterized in that, The method is executed by a terminal device, and the method includes: Send a first message, which is used to indicate the storage status of one or more artificial intelligence (AI) units deployed on the terminal device.

2. The method according to claim 1, characterized in that, The first information indicates the storage state of the one or more AI units via a single bit; or, The first information indicates the storage status of multiple AI units through a bit map, wherein one bit in the bit map corresponds to one AI unit among the multiple AI units.

3. The method according to claim 1 or 2, characterized in that, The AI ​​unit corresponds to the AI ​​model group deployed on the terminal device, and the AI ​​model group is a functional division based on the AI ​​models; or, The AI ​​unit corresponds to the AI ​​model deployed on the terminal device; or, The AI ​​unit corresponds to the parameter set of the first AI model, which is any AI model deployed on the terminal device.

4. The method according to any one of claims 1 to 3, characterized in that, In the case where the first information is used to indicate the storage status of multiple AI units, the storage status of the multiple AI units may be the same or different.

5. The method according to claim 4, characterized in that, The AI ​​unit in the first state is stored in the computing unit of the terminal device or a storage unit associated with the computing unit; or, The AI ​​unit in the second state is not stored in the computing unit of the terminal device or the storage unit associated with the computing unit.

6. The method according to claim 4 or 5, characterized in that, The storage space occupied by the AI ​​unit in the first state is less than or equal to the first threshold; and / or, In the first state, the storage space occupied by the AI ​​unit used to implement the first function is less than or equal to the second threshold.

7. The method according to claim 6, characterized in that, The first information is also used to indicate the first threshold and / or the second threshold.

8. The method according to any one of claims 1 to 7, characterized in that, The storage space occupied by the first AI unit is determined by the size of the storage space occupied by the first AI unit; or, The storage space occupied by the first AI unit is determined by the number of storage units occupied by the first AI unit, wherein the size of one of the storage units is predefined or preconfigured; Wherein, the first AI unit is any one of the one or more AI units.

9. The method according to any one of claims 1 to 8, characterized in that, The method further includes: Send a second message, which is used to update the storage state of the one or more AI units, or the second message is used to indicate the AI ​​unit whose storage state has changed.

10. The method according to any one of claims 1 to 9, characterized in that, If the first AI unit is in the second state, the first information is also used to indicate latency information, which is used to determine the latency unit corresponding to the first AI unit, and the first AI unit is any one of the one or more AI units.

11. The method according to claim 10, characterized in that, If the first AI unit is applied to an AI model group that implements the first function, the latency information is used to indicate the latency unit associated with the first function; or, If the first AI unit corresponds to the first AI model, the delay information is used to indicate the delay coefficient corresponding to the first AI model.

12. The method according to claim 11, characterized in that, The latency unit associated with the first AI model is the product of the latency coefficient corresponding to the first AI model and the storage space occupied by the first AI model.

13. The method according to any one of claims 10 to 12, characterized in that, The time delay unit is the time interval between the second time and the first time, where the first time is the moment when the first AI unit enters the first state.

14. The method according to claim 13, characterized in that, The second moment is the moment when the terminal device receives the first signaling, which is used to activate the first AI unit; or, The second moment is the moment when the terminal device receives the first channel, the first channel carrying a first signaling for activating the AI ​​unit; or, The second time is the time when the terminal device receives the first reference signal; or, The second moment is the moment when the terminal device receives the first reference resource.

15. The method according to any one of claims 1 to 14, characterized in that, The first information is carried in the capability information of the terminal device; or... The first information is carried in the auxiliary information of the terminal device.

16. A wireless communication method, characterized in that, The method is performed by a network device, and the method includes: Receive first information, which is used to indicate the storage status of one or more artificial intelligence (AI) units deployed on the terminal device.

17. The method according to claim 16, characterized in that, The first information indicates the storage state of the one or more AI units via a single bit; or, The first information indicates the storage status of multiple AI units through a bit map, wherein one bit in the bit map corresponds to one AI unit among the multiple AI units.

18. The method according to claim 16 or 17, characterized in that, The AI ​​unit corresponds to the AI ​​model group deployed on the terminal device, and the AI ​​model group is a functional division based on the AI ​​models; or, The AI ​​unit corresponds to the AI ​​model deployed on the terminal device; or, The AI ​​unit corresponds to the parameter set of the first AI model, which is any AI model deployed on the terminal device.

19. The method according to any one of claims 16 to 18, characterized in that, In the case where the first information is used to indicate the storage status of multiple AI units, the storage status of the multiple AI units may be the same or different.

20. The method according to claim 19, characterized in that, The AI ​​unit in the first state is stored in the computing unit of the terminal device or a storage unit associated with the computing unit; or, The AI ​​unit in the second state is not stored in the computing unit of the terminal device or the storage unit associated with the computing unit.

21. The method according to claim 19 or 20, characterized in that, The storage space occupied by the AI ​​unit in the first state is less than or equal to the first threshold; and / or, In the first state, the storage space occupied by the AI ​​unit used to implement the first function is less than or equal to the second threshold.

22. The method according to claim 21, characterized in that, The first information is also used to indicate the first threshold and / or the second threshold.

23. The method according to any one of claims 16 to 22, characterized in that, The storage space occupied by the first AI unit is determined by the size of the storage space occupied by the first AI unit; or, The storage space occupied by the first AI unit is determined by the number of storage units occupied by the first AI unit, wherein the size of one of the storage units is predefined or preconfigured; Wherein, the first AI unit is any one of the one or more AI units.

24. The method according to any one of claims 16 to 23, characterized in that, The method further includes: Receive second information, which is used to update the storage state of the one or more AI units, or the second information is used to indicate the AI ​​unit whose storage state has changed.

25. The method according to any one of claims 16 to 24, characterized in that, If the first AI unit is in the second state, the first information is also used to indicate latency information, which is used to determine the latency unit corresponding to the first AI unit, and the first AI unit is any one of the one or more AI units.

26. The method according to claim 25, characterized in that, If the first AI unit is applied to an AI model group that implements the first function, the latency information is used to indicate the latency unit associated with the first function; or, If the first AI unit corresponds to the first AI model, the delay information is used to indicate the delay coefficient corresponding to the first AI model.

27. The method according to claim 26, characterized in that, The latency unit associated with the first AI model is the product of the latency coefficient corresponding to the first AI model and the storage space occupied by the first AI model.

28. The method according to any one of claims 25 to 27, characterized in that, The time delay unit is the time interval between the second time and the first time, where the first time is the moment when the first AI unit enters the first state.

29. The method according to claim 28, characterized in that, The second moment is the moment when the terminal device receives the first signaling, which is used to activate the first AI unit; or, The second moment is the moment when the terminal device receives the first channel, the first channel carrying a first signaling for activating the AI ​​unit; or, The second time is the time when the terminal device receives the first reference signal; or, The second moment is the moment when the terminal device receives the first reference resource.

30. The method according to any one of claims 16 to 29, characterized in that, The first information is carried in the capability information of the terminal device; or... The first information is carried in the auxiliary information of the terminal device.

31. A terminal device, characterized in that, The terminal device includes: The sending module is used to send first information, which is used to indicate the storage status of one or more artificial intelligence (AI) units deployed in the terminal device.

32. The terminal device according to claim 31, characterized in that, The first information indicates the storage state of the one or more AI units via a single bit; or, The first information indicates the storage status of multiple AI units through a bit map, wherein one bit in the bit map corresponds to one AI unit among the multiple AI units.

33. The terminal device according to claim 31 or 32, characterized in that, The AI ​​unit corresponds to the AI ​​model group deployed on the terminal device, and the AI ​​model group is a functional division based on the AI ​​models; or, The AI ​​unit corresponds to the AI ​​model deployed on the terminal device; or, The AI ​​unit corresponds to the parameter set of the first AI model, which is any AI model deployed on the terminal device.

34. The terminal device according to any one of claims 31 to 33, characterized in that, In the case where the first information is used to indicate the storage status of multiple AI units, the storage status of the multiple AI units may be the same or different.

35. The terminal device according to claim 34, characterized in that, The AI ​​unit in the first state is stored in the computing unit of the terminal device or a storage unit associated with the computing unit; or, The AI ​​unit in the second state is not stored in the computing unit of the terminal device or the storage unit associated with the computing unit.

36. The terminal device according to claim 34 or 35, characterized in that, The storage space occupied by the AI ​​unit in the first state is less than or equal to the first threshold; and / or, In the first state, the storage space occupied by the AI ​​unit used to implement the first function is less than or equal to the second threshold.

37. The terminal device according to claim 36, characterized in that, The first information is also used to indicate the first threshold and / or the second threshold.

38. The terminal device according to any one of claims 31 to 37, characterized in that, The storage space occupied by the first AI unit is determined by the size of the storage space occupied by the first AI unit; or, The storage space occupied by the first AI unit is determined by the number of storage units occupied by the first AI unit, wherein the size of one of the storage units is predefined or preconfigured; Wherein, the first AI unit is any one of the one or more AI units.

39. The terminal device according to any one of claims 31 to 38, characterized in that, The sending module is further configured to send second information, which is used to update the storage state of the one or more AI units, or to indicate an AI unit whose storage state has changed.

40. The terminal device according to any one of claims 31 to 39, characterized in that, If the first AI unit is in the second state, the first information is also used to indicate latency information, which is used to determine the latency unit corresponding to the first AI unit, and the first AI unit is any one of the one or more AI units.

41. The terminal device according to claim 40, characterized in that, If the first AI unit is applied to an AI model group that implements the first function, the latency information is used to indicate the latency unit associated with the first function; or, If the first AI unit corresponds to the first AI model, the delay information is used to indicate the delay coefficient corresponding to the first AI model.

42. The terminal device according to claim 41, characterized in that, The latency unit associated with the first AI model is the product of the latency coefficient corresponding to the first AI model and the storage space occupied by the first AI model.

43. The terminal device according to any one of claims 40 to 42, characterized in that, The time delay unit is the time interval between the second time and the first time, where the first time is the moment when the first AI unit enters the first state.

44. The terminal device according to claim 43, characterized in that, The second moment is the moment when the terminal device receives the first signaling, which is used to activate the first AI unit; or, The second moment is the moment when the terminal device receives the first channel, the first channel carrying a first signaling for activating the AI ​​unit; or, The second time is the time when the terminal device receives the first reference signal; or, The second moment is the moment when the terminal device receives the first reference resource.

45. The terminal device according to any one of claims 31 to 44, characterized in that, The first information is carried in the capability information of the terminal device; or... The first information is carried in the auxiliary information of the terminal device.

46. ​​A network device, characterized in that, The network device includes: A receiving module is used to receive first information, which indicates the storage status of one or more artificial intelligence (AI) units deployed on the terminal device.

47. The network device according to claim 46, characterized in that, The first information indicates the storage state of the one or more AI units via a single bit; or, The first information indicates the storage status of multiple AI units through a bit map, wherein one bit in the bit map corresponds to one AI unit among the multiple AI units.

48. The network device according to claim 46 or 47, characterized in that, The AI ​​unit corresponds to the AI ​​model group deployed on the terminal device, and the AI ​​model group is a functional division based on the AI ​​models; or, The AI ​​unit corresponds to the AI ​​model deployed on the terminal device; or, The AI ​​unit corresponds to the parameter set of the first AI model, which is any AI model deployed on the terminal device.

49. The network device according to any one of claims 46 to 48, characterized in that, In the case where the first information is used to indicate the storage status of multiple AI units, the storage status of the multiple AI units may be the same or different.

50. The network device according to claim 49, characterized in that, The AI ​​unit in the first state is stored in the computing unit of the terminal device or a storage unit associated with the computing unit; or, The AI ​​unit in the second state is not stored in the computing unit of the terminal device or the storage unit associated with the computing unit.

51. The network device according to claim 49 or 50, characterized in that, The storage space occupied by the AI ​​unit in the first state is less than or equal to the first threshold; and / or, In the first state, the storage space occupied by the AI ​​unit used to implement the first function is less than or equal to the second threshold.

52. The network device according to claim 51, characterized in that, The first information is also used to indicate the first threshold and / or the second threshold.

53. The network device according to any one of claims 46 to 52, characterized in that, The storage space occupied by the first AI unit is determined by the size of the storage space occupied by the first AI unit; or, The storage space occupied by the first AI unit is determined by the number of storage units occupied by the first AI unit, wherein the size of one of the storage units is predefined or preconfigured; Wherein, the first AI unit is any one of the one or more AI units.

54. The network device according to any one of claims 46 to 53, characterized in that, The receiving module is further configured to receive second information, which is used to update the storage state of the one or more AI units, or to indicate an AI unit whose storage state has changed.

55. The network device according to any one of claims 46 to 54, characterized in that, If the first AI unit is in the second state, the first information is also used to indicate latency information, which is used to determine the latency unit corresponding to the first AI unit, and the first AI unit is any one of the one or more AI units.

56. The network device according to claim 55, characterized in that, If the first AI unit is applied to an AI model group that implements the first function, the latency information is used to indicate the latency unit associated with the first function; or, If the first AI unit corresponds to the first AI model, the delay information is used to indicate the delay coefficient corresponding to the first AI model.

57. The network device according to claim 56, characterized in that, The latency unit associated with the first AI model is the product of the latency coefficient corresponding to the first AI model and the storage space occupied by the first AI model.

58. The network device according to any one of claims 55 to 57, characterized in that, The time delay unit is the time interval between the second time and the first time, where the first time is the moment when the first AI unit enters the first state.

59. The network device according to claim 58, characterized in that, The second moment is the moment when the terminal device receives the first signaling, which is used to activate the first AI unit; or, The second moment is the moment when the terminal device receives the first channel, the first channel carrying a first signaling for activating the AI ​​unit; or, The second time is the time when the terminal device receives the first reference signal; or, The second moment is the moment when the terminal device receives the first reference resource.

60. The network device according to any one of claims 46 to 59, characterized in that, The first information is carried in the capability information of the terminal device; or... The first information is carried in the auxiliary information of the terminal device.

61. A terminal device, characterized in that, The terminal device includes a processor and a memory, the memory storing a computer program, and the processor executing the computer program to cause the terminal device to implement the method as described in any one of claims 1 to 15.

62. A network device, characterized in that, The network device includes a processor and a memory, the memory storing a computer program, the processor executing the computer program to cause the network device to perform the method as described in any one of claims 16 to 30.

63. A computer-readable storage medium, characterized in that, The storage medium stores a computer program that is executed by a processor to implement the method as described in any one of claims 1 to 15, or the method as described in any one of claims 16 to 30.

64. A chip, characterized in that, The chip includes programmable logic circuitry and / or program instructions, which, when the chip is running, are used to implement the method as described in any one of claims 1 to 15, or to implement the method as described in any one of claims 16 to 30.

65. A computer program product, characterized in that, The computer program product includes computer instructions stored in a computer-readable storage medium, which a processor reads from and executes to implement the method as claimed in any one of claims 1 to 15, or the method as claimed in any one of claims 16 to 30.