A data transmission method and related apparatus

Abstract

A data transmission method and related apparatus, the method comprising: obtaining an AI task processing request, the AI task processing request indicating a first task identifier of a first task and first input data of the first task, the first task comprising a training task and / or an inference task based on a first AI model; and sending the first input data to a first function entity in which the first AI model is deployed, the first function entity having a corresponding relationship with the first task identifier. The application implements the processing of AI tasks based on the corresponding relationship between the task identifier and the function entity, and different AI tasks can be executed by function entities providing different AI services, instead of being executed by a single network element (such as a network data analysis function (NWDAF) network element), thereby improving the execution efficiency of AI tasks through the task identifier of the AI task and the AI task allocation mechanism.

Description

Technical Field

[0001] This application relates to the field of data processing, and in particular to a data transmission method and related apparatus. Background Technology

[0002] The convergence of artificial intelligence (AI) and radio access network (RAN) is of great significance for revolutionizing network infrastructure, improving mobile network performance, and ensuring user experience.

[0003] In AI task processing solutions, the core network (CN) processes AI task-related data sent by user equipment (UE), specifically through the network data analytics function (NWDAF) within the core network. However, different AI tasks may require different resources, and a solution based on NWDAF for all AI tasks results in execution efficiency that fails to meet requirements. Summary of the Invention

[0004] This application provides a data transmission method and related apparatus for improving the execution efficiency of AI tasks.

[0005] The first aspect of this application provides a data transmission method, the method comprising: acquiring an artificial intelligence (AI) task processing request, wherein the AI ​​task processing request indicates a first task identifier and first input data of the first task, the first task including a training task and / or an inference task based on a first AI model; and sending the first input data to a first functional entity on which the first AI model is deployed, wherein there is a correspondence between the first functional entity and the first task identifier.

[0006] Based on the above scheme, there is a correspondence between the task identifier of an AI task and the functional entity. Input data for different AI tasks can be sent to the corresponding functional entity for processing based on the corresponding task identifier. Therefore, this application embodiment realizes the processing of AI tasks based on the correspondence between task identifiers and functional entities. Different AI tasks can be executed by functional entities that provide different AI services, instead of being executed by a single network element (such as the Network Data Analysis Function (NWDAF) network element). The execution efficiency of AI tasks is improved through the task identifier and AI task allocation mechanism.

[0007] It should be noted that the data transmission method provided in the first aspect of the embodiments of this application can be applied to both the terminal side and the network side, and the embodiments of this application do not limit the subject executing the method.

[0008] Among them, the terminal side includes, for example, the terminal or the communication / computing module in the terminal, or the circuit or chip in the terminal responsible for communication functions (such as modem chip, also known as baseband chip, or system on chip (SoC) chip containing modem core or system in package (SIP) chip), or the circuit or chip in the terminal responsible for computing and / or communication functions (such as graphics processing unit (GPU), artificial intelligence (AI) processor, or application specific integrated circuit (ASIC)), or the logic node, logic module or software that can realize all or part of the terminal functions.

[0009] Among them, network side, such as network side access network equipment, modules (such as circuits, chips or chip systems) in access network equipment, or logical nodes, logical modules or software that can realize all or part of the functions of access network equipment, or circuits or chips (such as GPUs, AI processors or ASICs) in access network equipment responsible for computing and / or communication functions, or network side core network (CN) equipment.

[0010] In this embodiment, the task identifier of the first task is used to uniquely identify the first task. The form of the task identifier is not limited in this application. The task identifier can be determined based on the process number corresponding to the task, or it can be determined based on the task type of the first task or the AI ​​service type corresponding to the first task. The first input data of the first task refers to the data to be processed corresponding to the first task, such as the input data for AI model inference, the training input data for AI model training, etc. The input data corresponding to different AI tasks is different, and the type of input data is not limited in this application.

[0011] In the above scheme, the step of obtaining the AI ​​task processing request can be understood as the method execution subject receiving the AI ​​task processing request from the outside. For example, if the execution subject is an access network device, the access network device receives the AI ​​task processing request sent by the terminal; if the execution subject is a core network element, the core network receives the AI ​​task processing request from the terminal sent by the access network device; or, it can also be understood as the internal implementation of the method execution subject. For example, if the execution subject is a terminal, the physical layer of the terminal receives the AI ​​task processing request sent by the upper layer (such as the application layer). In this embodiment, no limitation is made on the implementation method of obtaining the AI ​​task processing request.

[0012] It should be noted that the embodiments of this application do not limit the way in which the AI ​​task processing request indicates the first task identifier and the first input data of the first task. Specifically, it can be indicated directly, for example, by directly including the first task identifier and the first input data in the AI ​​task processing request; it can also be indicated indirectly, for example, by including other information for indicating the acquisition of the first task identifier and the first input data in the AI ​​task processing request; it can also be indicated by a combination of direct and indirect indication, for example, by directly including the first task identifier and other information for indicating the acquisition of the first input data in the AI ​​task processing request.

[0013] In this embodiment, the functional entities are associated with AI tasks. For example, AI tasks may include, but are not limited to, transceiver functions, channel estimation, positioning, and mobility management. The functional entities in this embodiment provide AI services. For example, AI services may include, but are not limited to, AI model inference services, AI model training services, and AI model training data support services. Furthermore, the functional entities in this embodiment may have pre-defined parameters related to the AI ​​tasks. For example, AI task-related parameters may include, but are not limited to, model parameters related to AI inference tasks and datasets related to AI model training tasks.

[0014] It should be noted that the functional entities in this application embodiment may be created and deployed to the network by an AI service provider. This application embodiment does not limit the form of the functional entities. The functional entities may be implemented by hardware, software, or a combination of software and hardware.

[0015] In one possible implementation of the first aspect, the method further includes: determining a first functional entity based on a first task identifier and a first mapping relationship, wherein the first mapping relationship includes the correspondence between the first task identifier and the first functional entity.

[0016] Based on the above scheme, this application embodiment pre-configures a first mapping relationship. The first mapping relationship includes the correspondence between task identifiers and functional entities. Different task identifiers can correspond to different functional entities or the same functional entity. Based on the first mapping relationship and the task identifier of the first task in the AI ​​task processing request, the functional entity used to process the first task can be determined. Therefore, this application embodiment provides an AI task allocation mechanism, where different AI tasks can be allocated to the functional entities corresponding to the task identifiers for execution, improving the execution efficiency of AI tasks.

[0017] In one possible implementation of the first aspect, the first functional entity is determined by the processing layer based on the first task identifier and the first mapping relationship.

[0018] The processing layer can be an existing data protocol layer or a newly added data protocol layer; this embodiment does not impose any limitations. The processing layer can be located between the application layer and the network access layer. The processing layer is pre-configured with a first mapping relationship, enabling the routing of target functional entities based on task identifiers.

[0019] In one possible implementation of the first aspect, the method further includes: obtaining an AI task control request, the AI ​​task control request indicating a first task identifier; and sending the AI ​​task control request to a first functional entity, the AI ​​task control request indicating one of: pausing the first task, restarting the first task, or terminating the first task.

[0020] Based on the above solution, the embodiments of this application can control the entire lifecycle of a task based on an AI task control request, such as controlling task pause, task restart, and task termination. By controlling the task, the needs of AI task processing can be responded to in a timely manner, further improving the execution efficiency of AI tasks.

[0021] It is understandable that task execution refers to the process by which a functional entity processes the input data of a task. The functional entity can respond to the process of processing the input data of a task in the form of a process. Correspondingly, the control of a task can be understood as the control of the process in the functional entity, such as the termination, restart and release of a process.

[0022] In one possible implementation of the first aspect, the method further includes: within a first time window, the first task identifier is also used to identify a second task, wherein the second task is a different task from the first task.

[0023] The first time window can be determined based on the processing time of the first input data. Within the first time window, the first task identifier can be reused, meaning it can be used to identify other AI tasks, such as the second task. Since AI tasks may require a long processing time during execution, such as high computational complexity or high inference latency, reusing the first task identifier within the processing time of the first input data can improve the utilization rate of task identifier resources.

[0024] In one possible implementation of the first aspect, obtaining the AI ​​task processing request includes: receiving an AI task processing request from a terminal, wherein the AI ​​task processing request is carried by a first transmission resource, and there is a correspondence between the first transmission resource and a first status identifier, wherein the first status identifier indicates the first status of the first task.

[0025] In the above scheme, the AI ​​task processing request is sent by the terminal. Specifically, the terminal sends the AI ​​task processing request based on the first transmission resource, and there is a correspondence between the first transmission resource and the first state identifier that can indicate the first state of the first task.

[0026] Understandably, the first state identifier is used to indicate the first state of the first task. The first state can be the task execution state of the first task. For example, if the first task is an AI model training task, the first state can include the AI ​​model training stage, the convergence status of the AI ​​model, the data consumption of the AI ​​model, the accuracy of the AI ​​model, etc. Optionally, the first state can also be associated with the task type of the first task, such as: the first task is an AI model training task, or an AI model inference task, etc.

[0027] Because different tasks in different states have different requirements for data transmission—for example, in the early stages of model training, the accuracy requirements for data interaction are not high, while in the later stages of model training, when the model is close to convergence, the accuracy of data transmission will increase accordingly—and because AI recognition tasks for autonomous driving have higher requirements for data transmission latency and reliability compared to image AI creation tasks, the embodiments of this application establish a correspondence between task state identifiers and transmission resources. Based on the task state identifier, the transmission resources used to carry AI task processing requests can be determined. This allows for the determination of different transmission resources based on different task states, with the transmission resources used to carry the data corresponding to the task. This provides suitable transmission resources for AI tasks in different states, thereby improving the execution efficiency of AI tasks.

[0028] In one possible implementation of the first aspect, the method further includes: sending first output data of the first task to the terminal based on the first transmission resource; or, sending the first output data to the terminal based on the second transmission resource, wherein there is a correspondence between the second transmission resource and the second status identifier, and the second status identifier is used to indicate the second status of the first task.

[0029] In the above scheme, after the first functional entity receives the first input data and performs corresponding processing, it obtains the first output data of the first task and sends it to the network side. Then, the network side can send the first output data to the terminal based on the first transmission resources. The first transmission resources are the resources used by the terminal to send the AI ​​task processing request. Thus, the network side directly sends the first output data to the terminal according to the resources used by the terminal to send the AI ​​task processing request, without having to additionally determine the transmission resources used for the first output data, thus saving the network side's data transmission overhead.

[0030] In another alternative approach, the network side can send the first output data to the terminal device based on the second transmission resource. There is a correspondence between the second transmission resource and a second status identifier, whereby the second status identifier indicates the second status of the first task. That is, the network side needs to determine the second status identifier of the first task and, based on the second status identifier, determine the second transmission resource used to carry the first output data.

[0031] Understandably, after processing the first input data, the functional entity obtains the first output data. The state of the first task may change, for example, the convergence status of the corresponding model may change. Therefore, the network side determines the transmission resources based on the task's state identifier, which makes the second transmission resources used to carry the first output data more adaptable to the current state of the first task, providing suitable transmission resources for the AI ​​task and thus improving the execution efficiency of the AI ​​task.

[0032] It should be noted that both the first state and the second state are states of the first task. The first state and the second state refer to the states of the first task in different task lifecycles. The first state is the state of the first task when the terminal sends an AI task processing request to the network side, and the second state is the state of the first task when the network side sends the first output data to the terminal. The first state and the second state may be different or the same; correspondingly, the first state identifier and the second state identifier may be the same or different.

[0033] The second aspect of this application provides a data transmission method that can be applied to the terminal side, such as a terminal or a communication / computing module in the terminal, or a circuit or chip in the terminal responsible for communication functions (such as a modem chip, also known as a baseband chip, or a system-on-chip (SoC) chip containing a modem core or a system-in-package (SIP) chip), or a circuit or chip in the terminal responsible for computing and / or communication functions (such as a graphics processing unit (GPU), an artificial intelligence (AI) processor, or an application-specific integrated circuit (ASIC)), or a logical node, logical module, or software that can implement all or part of the terminal functions.

[0034] Taking the application of this method to a terminal as an example, in this method, an AI task processing request is sent, which indicates the first task identifier of the first task and the first input data of the first task. The first task includes a training task and / or an inference task based on a first AI model. The first output data of the first task is received from a first functional entity, and there is a correspondence between the first functional entity and the first task identifier.

[0035] Based on the above scheme, there is a correspondence between the task identifier of an AI task and a functional entity. Input data for different AI tasks can be sent to the corresponding functional entity for processing based on the corresponding task identifier. Therefore, this application embodiment realizes the processing of AI tasks based on the correspondence between task identifiers and functional entities. Different AI tasks can be executed by functional entities that provide different AI services, instead of being executed by a single network element (such as the Network Data Analysis Function (NWDAF) network element). This application embodiment improves the execution efficiency of AI tasks through the task identifier and AI task allocation mechanism.

[0036] In this embodiment, the task identifier of the first task is used to uniquely identify the first task. The form of the task identifier is not limited in this application. The task identifier can be determined based on the process number corresponding to the task, or it can be determined based on the task type of the first task or the AI ​​service type corresponding to the first task. The first input data of the first task refers to the data to be processed corresponding to the first task, such as the input data for AI model inference, the training input data for AI model training, etc. The input data corresponding to different AI tasks is different, and the type of input data is not limited in this application.

[0037] It should be noted that the embodiments of this application do not limit the way in which the AI ​​task processing request indicates the first task identifier and the first input data of the first task. Specifically, it can be indicated directly, for example, by directly including the first task identifier and the first input data in the AI ​​task processing request; it can also be indicated indirectly, for example, by including other information for indicating the acquisition of the first task identifier and the first input data in the AI ​​task processing request; it can also be indicated by a combination of direct and indirect indication, for example, by directly including the first task identifier and other information for indicating the acquisition of the first input data in the AI ​​task processing request.

[0038] In this embodiment, the functional entities are associated with AI tasks. For example, AI tasks may include, but are not limited to, transceiver functions, channel estimation, positioning, and mobility management. The functional entities in this embodiment provide AI services. For example, AI services may include, but are not limited to, AI model inference services, AI model training services, and AI model training data support services. Furthermore, the functional entities in this embodiment may have pre-defined parameters related to the AI ​​tasks. For example, AI task-related parameters may include, but are not limited to, model parameters related to AI inference tasks and datasets related to AI model training tasks.

[0039] It should be noted that the functional entities in this application embodiment may be created and deployed to the network by an AI service provider. This application embodiment does not limit the form of the functional entities. The functional entities may be implemented by hardware, software, or a combination of software and hardware.

[0040] In one possible implementation of the second aspect, the method further includes: sending an AI task control request to a first functional entity, the AI ​​task control request indicating a first task identifier, the AI ​​task control request being used to indicate one of: pausing the first task, restarting the first task, and terminating the first task.

[0041] Based on the above solution, the embodiments of this application can control the entire lifecycle of a task based on an AI task control request, such as controlling task pause, task restart, and task termination. By controlling the task, the needs of AI task processing can be responded to in a timely manner, further improving the execution efficiency of AI tasks.

[0042] It is understandable that task execution refers to the process by which a functional entity processes the input data of a task. The functional entity can respond to the process of processing the input data of a task in the form of a process. Correspondingly, the control of a task can be understood as the control of the process in the functional entity, such as the termination, restart and release of a process.

[0043] In one possible implementation of the second aspect, within the first time window, the first task identifier is also used to identify the second task, which is a different task from the first task.

[0044] The first time window can be determined based on the processing time of the first input data. Within the first time window, the first task identifier can be reused, meaning it can be used to identify other AI tasks, such as the second task. Since AI tasks may require a long processing time during execution, such as high computational complexity or high inference latency, reusing the first task identifier within the processing time of the first input data can improve the utilization rate of task identifier resources.

[0045] In one possible implementation of the second aspect, the method further includes: obtaining a first state identifier, the first state identifier being used to indicate a first state of the first task; and sending an AI task processing request, including: sending the AI ​​task processing request based on a first transmission resource, wherein there is a correspondence between the first transmission resource and the first state identifier.

[0046] In the above scheme, when the terminal determines the first transmission resource to carry the AI ​​task processing request, it needs to obtain the first status identifier of the first task and determine the first transmission resource based on the first status identifier. There is a correspondence between the first transmission resource and the first status identifier.

[0047] Understandably, the first state identifier is used to indicate the first state of the first task. The first state can be the task execution state of the first task. For example, if the first task is an AI model training task, the first state can include the AI ​​model training stage, the convergence status of the AI ​​model, the data consumption of the AI ​​model, the accuracy of the AI ​​model, etc. Optionally, the first state can also be associated with the task type of the first task, such as: the first task is an AI model training task, or an AI model inference task, etc.

[0048] Because different tasks in different states have different requirements for data transmission—for example, in the early stages of model training, the accuracy requirements for data interaction are not high, while in the later stages of model training, when the model is close to convergence, the accuracy of data transmission will increase accordingly—and because AI recognition tasks for autonomous driving have higher requirements for data transmission latency and reliability compared to image AI creation tasks, the embodiments of this application establish a correspondence between task state identifiers and transmission resources. Based on the task state identifier, the transmission resources used to carry AI task processing requests can be determined. This allows for the determination of different transmission resources based on different task states, with the transmission resources used to carry the data corresponding to the task. This provides suitable transmission resources for AI tasks in different states, thereby improving the execution efficiency of AI tasks.

[0049] In one possible implementation of the first aspect, the first output data is carried by the first transmission resource; or, the first output data is carried by a second transmission resource, wherein there is a correspondence between the second transmission resource and a second status identifier, and the second status identifier is used to indicate the second status of the first task.

[0050] In the above scheme, the terminal can receive the first output data of the first task sent by the network side. Specifically, in one optional approach, the terminal can receive the first output data sent by the network side based on a first transmission resource, which is the resource used by the terminal to send the AI ​​task processing request. Thus, the network side directly sends the first output data to the terminal according to the resource used by the terminal to send the AI ​​task processing request, without needing to additionally determine the transmission resource used for the first output data, saving the network overhead when sending the first output data.

[0051] In another alternative approach, the terminal can accept first output data sent by the network side based on a second transmission resource. There is a correspondence between the second transmission resource and a second state identifier, with the second state identifier indicating the second state of the first task. That is, the network side needs to re-determine the second state identifier of the first task and determine the second transmission resource to carry the first output data based on it. It is understood that after a functional entity processes the first input data to obtain the first output data, the state of the first task may change; for example, the convergence status of the corresponding model may change. Therefore, by determining the transmission resource based on the task's state identifier, the network side can make the second transmission resource used to carry the first output data more adaptable to the current state of the first task, providing suitable transmission resources for the AI ​​task and thus improving the execution efficiency of the AI ​​task.

[0052] It should be noted that both the first state and the second state are states of the first task. The first state and the second state refer to the states of the first task in different task lifecycles. The first state is the state of the first task when the terminal sends an AI task processing request to the network side, and the second state is the state of the first task when the network side sends the first output data to the terminal. The first state and the second state may be different or the same; correspondingly, the first state identifier and the second state identifier may be the same or different.

[0053] Thirdly, this application provides a communication device that has the functions of the first aspect described above. For example, the communication device includes modules, units, or means that perform the operations involved in the first aspect. These modules, units, or means can be implemented by software, hardware, or a combination of software and hardware.

[0054] Fourthly, this application provides a communication device that has the functions of the second aspect above. For example, the communication device includes modules, units, or means that perform the operations involved in the second aspect above. These modules, units, or means can be implemented by software, hardware, or a combination of software and hardware.

[0055] Fifthly, this application provides a communication device including an interface circuit and one or more processors. The one or more processors are coupled to a memory. The memory stores part or all of the necessary computer program or instructions for implementing the functions described in the second aspect above. The one or more processors can execute the computer program or instructions, causing the communication device to implement the methods in any possible design or implementation of the second aspect above when executed. The interface circuit is used to implement the communication functions within the communication device and / or the communication functions between the communication device and other devices or components.

[0056] In one possible design, the processor is used to communicate with other devices or components through the interface circuit.

[0057] In one possible design, the communication device may also include the memory.

[0058] The aforementioned communication device may be a terminal, or a communication / computing module within a terminal, or a chip within a terminal responsible for communication functions such as a modem chip (also known as a baseband chip) or a SoC or SIP chip containing a modem module, or a circuit or chip within a terminal responsible for computing and / or communication functions (such as a GPU, AI processor, or ASIC), or a logical node or logical module capable of implementing all or part of the terminal's functions. The aforementioned communication device may also be an access network device, or a module within an access network device (e.g., a circuit, chip, or chip system), or a circuit or chip within an access network device responsible for computing and / or communication functions (such as a GPU, AI processor, or ASIC), or a logical node or logical module capable of implementing all or part of the access network device's functions.

[0059] Sixthly, this application provides a communication device including an interface circuit and one or more processors. The one or more processors are coupled to a memory. The memory stores part or all of the necessary computer program or instructions for implementing the functions described in the first aspect above. The one or more processors are executable to carry out the computer program or instructions, causing the communication device to implement the methods in any possible design or implementation of the first aspect above. The interface circuit is used to implement the communication functions within the communication device and / or the communication functions between the communication device and other devices or components.

[0060] In one possible design, the processor is used to communicate with other devices or components through the interface circuit.

[0061] In one possible design, the communication device may also include the memory.

[0062] The aforementioned communication device may be a terminal, or a communication / computing module in a terminal, or a chip in a terminal responsible for communication functions such as a modem chip (also known as a baseband chip) or a SoC or SIP chip containing a modem module, or a circuit or chip in a terminal responsible for computing and / or communication functions (such as a GPU, AI processor, or ASIC), or a logical node or logical module capable of implementing all or part of the terminal functions.

[0063] In a seventh aspect, this application provides a communication system, which includes the communication device of the fifth aspect and / or the communication device of the sixth aspect described above.

[0064] Eighthly, this application provides a computer-readable storage medium storing computer-readable instructions that, when read and executed by a computer, cause the computer to perform any of the possible designs in the first to second aspects described above.

[0065] Ninthly, this application provides a computer program product that, when read and executed by a computer, causes the computer to perform any of the possible designs in the first to second aspects described above.

[0066] The technical effects of any of the design methods in aspects three through nine can be found in the technical effects of the different design methods in aspects one through two above, and will not be repeated here. Attached Figure Description

[0067] Figures 1-5 A schematic diagram of the communication system provided in the embodiments of this application;

[0068] Figure 6 and Figure 7 A flowchart illustrating the data transmission method provided in an embodiment of this application;

[0069] Figure 8 and Figure 9 A schematic diagram illustrating the method for determining transmission resources provided in an embodiment of this application;

[0070] Figure 10 This application provides a schematic diagram of the data structure of a protocol data unit corresponding to an AI task processing request, as shown in an embodiment of the present application.

[0071] Figure 11 This application provides a schematic diagram of a delay information transmission process.

[0072] Figure 12 This is a schematic diagram of the structure of a communication device provided in an embodiment of this application;

[0073] Figure 13This is a schematic diagram of a terminal structure provided in an embodiment of this application. Detailed Implementation

[0074] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms are interchangeable where appropriate; this is merely a way of distinguishing objects with the same attributes in the embodiments of this application. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, so that a process, method, system, product, or apparatus that comprises a series of elements is not necessarily limited to those elements, but may include other elements not explicitly listed or inherent to those processes, methods, products, or apparatuses.

[0075] The following explanations of some terms used in the embodiments of this application are provided to facilitate understanding by those skilled in the art.

[0076] (1) Terminal device: can be a wireless terminal device that can receive network device scheduling and instruction information. The wireless terminal device can be a device that provides voice and / or data connectivity to the user, or a handheld device with wireless connection function, or other processing device connected to a wireless modem.

[0077] Terminal devices can be various communication kits with wireless communication capabilities (the kit may include, for example, antennas, power supply modules, cables, and Wi-Fi modules). Terminal devices can also be communication modules with satellite communication capabilities, satellite phones or components thereof, and very small aperture terminals (VSATs). Terminal devices can be mobile terminal devices, such as mobile phones (or "cellular" phones), computers, and data cards. For example, they can be portable, pocket-sized, handheld, computer-embedded, or vehicle-mounted mobile devices that exchange voice and / or data with a wireless access network. Examples include personal communication service (PCS) phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (PDAs), tablets, and computers with wireless transceiver capabilities. Wireless terminal equipment can also be referred to as a system, subscriber unit, subscriber station, mobile station, mobile station (MS), remote station, access point (AP), remote terminal, access terminal, user terminal, user agent, subscriber station (SS), customer premises equipment (CPE), terminal, user equipment (UE), mobile terminal (MT), drone, etc. Terminal equipment can also be wearable devices and next-generation communication systems, such as terminal equipment in 6G communication systems or terminal equipment in future evolved public land mobile networks (PLMNs). Of course, in this application, terminal equipment can also refer to chips, modems, system-on-a-chip (SoC), or communication platforms that may include radio frequency (RF) components, etc., that are primarily responsible for related communication functions.

[0078] In this embodiment of the application, the terminal device can also be a device with an artificial intelligence (AI) module, which can provide AI services to the terminal device.

[0079] (2) Network equipment: This can be equipment in a wireless network. For example, network equipment can be a RAN node (or device) that connects terminal devices to the wireless network, and can also be called a base station. Currently, some examples of RAN equipment include: base station, evolved NodeB (eNodeB), gNB (gNodeB) in 5G communication systems, transmission reception point (TRP), evolved Node B (eNB), radio network controller (RNC), Node B (NB), home base station (e.g., home-evolved Node B, or home Node B, HNB), base band unit (BBU), or wireless fidelity (Wi-Fi) access point (AP), etc. In addition, in a network structure, network equipment can include centralized unit (CU) nodes, distributed unit (DU) nodes, or RAN equipment including CU nodes and DU nodes.

[0080] Optionally, the RAN node can also be a macro base station, micro base station, indoor station, relay node, donor node, or a wireless controller in a cloud radio access network (CRAN) scenario. The RAN node can also be a server, wearable device, vehicle, or in-vehicle equipment. For example, the access network equipment in vehicle-to-everything (V2X) technology can be a roadside unit (RSU). All or part of the functions of the RAN node in this application can also be implemented through software functions running on hardware, or through virtualization functions instantiated on a platform (e.g., a cloud platform). The RAN node can also be equipped with communication modules, circuits, or chips that perform corresponding communication functions. The RAN node can also be configured with program instructions for performing corresponding communication functions and corresponding program instructions. The RAN node in this application can also be a logical node, logical module, or software capable of implementing all or part of the access node functions, or a circuit or chip (such as a graphics processing unit (GPU), artificial intelligence (AI) processor, or application-specific integrated circuit (ASIC)) responsible for computing and / or communication functions in the access node.

[0081] In another possible scenario, multiple RAN nodes collaborate to assist the terminal in achieving wireless access, with different RAN nodes each implementing a portion of the base station's functions. For example, RAN nodes can be central units (CUs), distributed units (DUs), CU-control plane (CPs), CU-user plane (UPs), or radio units (RUs), etc. CUs and DUs can be set up separately or included in the same network element, such as a baseband unit (BBU). RUs can be included in radio frequency equipment or radio frequency units, such as remote radio units (RRUs), active antenna units (AAUs), or remote radio heads (RRHs).

[0082] In different systems, CU (or CU-CP and CU-UP), DU, or RU may have different names, but those skilled in the art will understand their meaning. For example, in an open access network (open RAN, O-RAN, or ORAN) system, CU can also be called O-CU (open CU), DU can also be called O-DU, CU-CP can also be called O-CU-CP, CU-UP can also be called O-CU-UP, and RU can also be called O-RU. For ease of description, this application uses CU, CU-CP, CU-UP, DU, and RU as examples. Any of the units among CU (or CU-CP, CU-UP), DU, and RU in this application can be implemented through software modules, hardware modules, or a combination of software modules and hardware modules.

[0083] Communication between access network devices and terminal devices follows a specific protocol layer structure. This protocol layer may include a control plane protocol layer and a user plane protocol layer. The control plane protocol layer may include at least one of the following: radio resource control (RRC) layer, packet data convergence protocol (PDCP) layer, radio link control (RLC) layer, media access control (MAC) layer, or physical (PHY) layer, etc. The user plane protocol layer may include at least one of the following: service data adaptation protocol (SDAP) layer, PDCP layer, RLC layer, MAC layer, or PHY layer, etc.

[0084] The correspondence between network elements and their achievable protocol layer functions in the ORAN system can be found in Table 1 below.

[0085] Table 1

[0086] ORAN network elements 3GPP protocol layer functions O-CU-CP RRC+PDCP-Control Plane (PDCP-C) O-CU-UP SDAP+PDCP - User Plane (PDCP-U) O-DU RLC+MAC+PHY-high O-RU PHY-low

[0087] Network devices can be other devices that provide wireless communication functions for terminal devices. The embodiments of this application do not limit the specific technology or form of the network device. For ease of description, the embodiments of this application are not limited.

[0088] Network equipment may also include core network equipment, such as the Mobility Management Entity (MME), Home Subscriber Server (HSS), Serving Gateway (S-GW), Policy and Charging Rules Function (PCRF), and Public Data Network Gateway (PDN Gateway, P-GW) in 4th generation (4G) networks; and access and mobility management function (AMF), user plane function (UPF), or session management function (SMF) in 5G networks. Furthermore, this core network equipment may also include other core network equipment in 5G networks and next-generation networks of 5G networks.

[0089] In this embodiment of the application, the network device mentioned above can also be a network node with artificial intelligence (AI) capabilities, which can provide AI services to terminals or other network devices. For example, it can be an AI node, computing power node, RAN node with AI capabilities, core network element with AI capabilities, etc. on the network side (access network or core network).

[0090] In this application embodiment, the device for implementing the function of the network device can be the network device itself, or it can be a device capable of supporting the network device in implementing that function, such as a chip system, which can be installed in the network device. In the technical solutions provided in this application embodiment, the example of a network device being used to implement the function of the network device is used to describe the technical solutions provided in this application embodiment.

[0091] (3) Configuration and Pre-configuration: In this application, both configuration and pre-configuration are used. Configuration refers to the network device sending configuration information or parameter values ​​of some parameters to the terminal device through messages or signaling, so that the terminal device can determine the communication parameters or resources during transmission based on these values ​​or information. Pre-configuration is similar to configuration; it can be parameter information or parameter values ​​that the network device and the terminal device have negotiated in advance, or it can be parameter information or parameter values ​​that the network device or the terminal device uses as specified by the standard protocol, or it can be parameter information or parameter values ​​that are pre-stored in the network device or the terminal device. This application does not limit this.

[0092] Furthermore, these values ​​and parameters can be changed or updated.

[0093] (4) The terms "system" and "network" in the embodiments of this application can be used interchangeably. "At least one" means one or more, and "more" means two or more. "And / or" describes the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, or B exists alone, where A and B can be singular or plural. The character " / " generally indicates that the related objects before and after are in an "or" relationship. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, "at least one of A, B and C" includes A, B, C, AB, AC, BC or ABC. And, unless otherwise specified, the ordinal numbers such as "first" and "second" mentioned in the embodiments of this application are used to distinguish multiple objects and are not used to limit the order, sequence, priority or importance of multiple objects.

[0094] (5) In the embodiments of this application, "send" and "receive" indicate the direction of signal transmission. For example, "send information to XX" can be understood as the destination of the information being XX, which may include sending directly through the air interface or sending indirectly through the air interface by other units or modules. "Receive information from YY" can be understood as the source of the information being YY, which may include receiving directly from YY through the air interface or receiving indirectly from YY through the air interface by other units or modules. "Send" can also be understood as the "output" of the chip interface, and "receive" can also be understood as the "input" of the chip interface.

[0095] In other words, sending and receiving can occur between devices, such as between network devices and terminal devices, or within a device, such as between components, modules, chips, software modules, or hardware modules within the device via buses, wiring, or interfaces.

[0096] It is understandable that information may undergo necessary processing, such as encoding and modulation, between the source and destination, but the destination can understand the valid information from the source. Similar statements in this application can be interpreted in a similar way and will not be elaborated further.

[0097] (6) In the embodiments of this application, "instruction" may include direct instruction and indirect instruction, as well as explicit instruction and implicit instruction. The information indicated by a certain piece of information (as described below, the instruction information) is called the information to be instructed. In the specific implementation process, there are many ways to indicate the information to be instructed, such as, but not limited to, directly indicating the information to be instructed, such as the information to be instructed itself or its index. It can also indirectly indicate the information to be instructed by indicating other information, where there is an association between the other information and the information to be instructed; or it can only indicate a part of the information to be instructed, while the other parts of the information to be instructed are known or pre-agreed upon. For example, the instruction can be implemented by using a pre-agreed (e.g., protocol predefined) arrangement order of various information, thereby reducing the instruction overhead to a certain extent. This application does not limit the specific method of instruction. It is understood that for the sender of the instruction information, the instruction information can be used to indicate the information to be instructed, and for the receiver of the instruction information, the instruction information can be used to determine the information to be instructed.

[0098] In this application, unless otherwise specified, the same or similar parts between the various embodiments can be referred to each other. In the various embodiments of this application, and the various methods / designs / implementations within each embodiment, unless otherwise specified or logically conflicting, the terminology and / or descriptions between different embodiments and between the various methods / designs / implementations within each embodiment are consistent and can be mutually referenced. The technical features in different embodiments and the various methods / designs / implementations within each embodiment can be combined to form new embodiments, methods, or implementations based on their inherent logical relationships. The following descriptions of the embodiments of this application do not constitute a limitation on the scope of protection of this application.

[0099] The technical solutions of this application can be applied to various data processing communication systems, such as code division multiple access (CDMA), time division multiple access (TDMA), frequency division multiple access (FDMA), orthogonal frequency-division multiple access (OFDMA), single-carrier frequency division multiple access (SC-FDMA), and other systems. Optionally, this application can be applied to long term evolution (LTE) systems, new radio (NR) systems, or new radio vehicle-to-everything (NR V2X) systems; it can also be applied to systems with hybrid LTE and 5G networks; or device-to-device (D2D) communication systems, machine-to-machine (M2M) communication systems, Internet of Things (IoT) systems, or drone communication systems; or communication systems supporting multiple wireless technologies, such as those supporting LTE and NR technologies; or non-terrestrial communication systems, such as satellite communication systems, high-altitude communication platforms, etc. Additionally, this communication system can also be applied to narrowband Internet of Things (NB-IoT) systems, or other communication systems.

[0100] Please see Figure 1 This is a schematic diagram of the architecture of the communication system 1000 used in an embodiment of this application. Figure 1 As shown, the communication system includes a radio access network (RAN) 100 and a core network 200. Optionally, the communication system 10 may also include an Internet 300. The RAN 100 includes at least one RAN node (e.g., ...). Figure 1 110a and 110b, collectively referred to as 110, may also include at least one terminal (such as...). Figure 1 RAN100, denoted as RAN100, comprises RAN nodes 120a-120j, collectively referred to as RAN120. RAN100 may also include other RAN nodes, such as wireless relay equipment and / or wireless backhaul equipment. Figure 1(Not shown in the image). Terminal 120 connects wirelessly to RAN node 110, and RAN node 110 connects wirelessly or via a wired connection to core network 200. The core network equipment in core network 200 and RAN node 110 in RAN 100 can be independent physical devices, or they can be the same physical device integrating the logical functions of core network equipment and RAN nodes. Terminals can connect to each other, and RAN nodes can connect to each other, via wired or wireless connections.

[0101] RAN 100 can be a cellular system related to the 3rd Generation Partnership Project (3GPP), such as 4G, 5G mobile communication systems, or future-oriented evolution systems. RAN 100 can also be an open RAN (O-RAN or ORAN), a cloud RAN (CRAN), a virtualized RAN (vRAN), an artificial intelligence radio access network (AI RAN), or a wireless fidelity (WiFi) system. RAN 100 can also be a communication system that integrates two or more of the above systems.

[0102] RAN node 110, sometimes also referred to as access network equipment, RAN entity, or access node, constitutes part of the communication system and is used to help terminals achieve wireless access. Multiple RAN nodes 110 in communication system 10 can be of the same type or different types. In some scenarios, the roles of RAN node 110 and terminal 120 are relative, for example... Figure 1 Network element 120i can be a helicopter or a drone, and it can be configured as a mobile base station. For terminals 120j that access RAN 100 through network element 120i, network element 120i is a base station; however, for base station 110a, network element 120i is a terminal. RAN node 110 and terminal 120 are sometimes referred to as communication devices, for example... Figure 1 Network elements 110a and 110b can be understood as communication devices with base station functions, while network elements 120a-120j can be understood as communication devices with terminal functions.

[0103] It should be noted that the technical solutions of the embodiments of this application are applicable to terrestrial communication systems, or to satellite communication systems, or to communication systems that integrate terrestrial and satellite communication. This communication system can also be referred to as a non-terrestrial network (NTN) communication system. For example, Figure 1 RAN100 in the above can include a ground base station, wherein the ground base station can include a TN cell (i.e., the signal of the TN cell can be transmitted and received through the ground base station); and, Figure 1 RAN100 can also include non-terrestrial base stations. Taking a satellite as an example, the satellite can include NTN cells (i.e., the signals of the NTN cells can be transmitted and received via the satellite). The terrestrial communication system can be, for example, a long term evolution (LTE) system, a 5G communication system, a new radio (NR) system, or a communication system that is the next step in the development of 5G communication systems, etc., without limitation here.

[0104] To support artificial intelligence (AI) technology in wireless networks, AI nodes may also be introduced into the network.

[0105] AI nodes can be deployed in one or more of the following locations within the communication system: access network nodes (RAN nodes), terminal devices, or core network devices. Alternatively, AI nodes can be deployed independently, for example, in a location other than any of the aforementioned devices, such as in the host or cloud server of an over-the-top (OTT) system. AI nodes can communicate with other devices in the communication system, which can be one or more of the following: network devices, terminal devices, or core network elements.

[0106] It is understood that this application does not limit the number of AI nodes. For example, when there are multiple AI nodes, these nodes can be divided based on function, such as different AI nodes being responsible for different functions.

[0107] It can also be understood that AI nodes can be independent devices, or they can be integrated into the same device to achieve different functions. Alternatively, they can be network elements in hardware devices, software functions running on dedicated hardware, or virtualization functions instantiated on a platform (e.g., a cloud platform). This application does not limit the specific form of the aforementioned AI nodes.

[0108] AI nodes can be AI network elements, AI modules, or AI functional entities.

[0109] Please see Figure 2 This is a schematic diagram of the architecture of another communication system used in an embodiment of this application. Figure 2 As shown, network elements in this communication system are connected via interfaces (e.g., NG, Xn) or air interfaces. These network element nodes, such as core network equipment, access network nodes (RAN nodes), terminals, or one or more devices in operations administration and maintenance (OAM), are equipped with one or more AI modules (for clarity, ...). Figure 2 (Only one is shown in the image). An access network node can be a single RAN node or can include multiple RAN nodes, such as CU and DU. CU and / or DU can also be configured with one or more AI modules. CU can also be split into CU-CP and CU-UP, with one or more AI modules configured in CU-CP and / or CU-UP.

[0110] AI modules are used to implement corresponding AI functions. AI modules deployed in different network elements can be the same or different. The models of AI modules can achieve different functions depending on the parameter configurations. The models of AI modules can be configured based on one or more of the following parameters: structural parameters (e.g., at least one of the following: number of neural network layers, neural network width, inter-layer connections, neuron weights, neuron activation function, or biases in the activation function), input parameters (e.g., the type and / or dimension of the input parameters), or output parameters (e.g., the type and / or dimension of the output parameters). The biases in the activation function can also be referred to as the biases of the neural network.

[0111] In one example, the neural network mentioned above could be a deep neural network (DNN), a convolutional neural network (CNN), a recurrent neural network (RNN), or a generative adversarial network (GAN).

[0112] Deep Neural Networks (DNNs) are artificial neural network architectures with multiple layers of nonlinear transformation units stacked in a hierarchical structure to form deep computational models. Compared to shallow neural networks, deep neural networks have more hidden layers, allowing the network model to capture more complex data structures and higher-level abstract features.

[0113] A CNN is a deep neural network with a convolutional structure. A CNN contains a feature extractor consisting of convolutional layers and subsampling layers. This feature extractor can be viewed as a filter, and the convolution process can be seen as performing convolution between a trainable filter and an input image or a convolutional feature map.

[0114] RNN is a type of recursive neural network that takes sequence data as input, recursively moves along the direction of sequence evolution, and connects all nodes (recurrent units) in a chain-like manner.

[0115] GAN is a deep learning model. It consists of a generator and a discriminator, and is trained through adversarial learning. Its purpose is to estimate the potential distribution of data samples and generate new data samples.

[0116] An AI module can have one or more models. A model can infer an output, which includes one or more parameters. The learning, training, or inference processes of different models can be deployed on different nodes or devices, or they can be deployed on the same node or device.

[0117] Please see Figure 3 This is a schematic diagram of the architecture of another communication system used in an embodiment of this application. Figure 3 As shown, this communication system includes a RAN intelligent controller (RIC). For example, the RIC could be... Figure 2 The AI ​​module shown is used to implement AI-related functions. RICs include near-real-time RICs (near-RT RICs) and non-real-time RICs (non-RT RICs). Non-real-time RICs primarily process non-real-time information, such as data that is not sensitive to latency, with latency in the order of seconds. Real-time RICs primarily process near-real-time information, such as data that is relatively sensitive to latency, with latency in the order of tens of milliseconds.

[0118] Near real-time (NRT) RICs are used for model training and inference. For example, they are used to train AI models and then use those models for inference. NRT RICs can obtain network-side and / or terminal-side information from RAN nodes (e.g., CUs, CU-CPs, CU-UPs, DUs, compute nodes, and / or RUs) and / or terminals. This information can be used as training data or inference data. NRT RICs can deliver inference results to RAN nodes and / or terminals. Inference results can be exchanged between CUs and DUs, and / or between DUs and RUs. For example, a NRT RIC delivers an inference result to a DU, which then forwards it to an RU.

[0119] Non-real-time RICs are also used for model training and inference. For example, they are used to train AI models and then use those models for inference. Non-real-time RICs can obtain network-side and / or terminal-side information from RAN nodes (e.g., CUs, CU-CPs, CU-UPs, DUs, compute nodes, and / or RUs) and / or terminals. This information can be used as training data or inference data, and the inference results can be delivered to RAN nodes and / or terminals. Inference results can be exchanged between CUs and DUs, and / or between DUs and RUs; for example, a non-real-time RIC delivers inference results to a DU, which then forwards them to an RU.

[0120] Near real-time RICs and non-real-time RICs can also be configured as separate network elements. Near real-time RICs and non-real-time RICs can also be part of other devices. For example, near real-time RICs can be set in RAN nodes (e.g., CU, DU, compute nodes), while non-real-time RICs can be set in OAM, cloud servers, core network devices, or other network devices.

[0121] Reference Figure 4The diagram illustrates another communication system that can be used to implement the data transmission method provided in the embodiments of this application. The user equipment (UE) includes an access network protocol (AN protocol) layer connected to the AN protocol layer in the access network node (RAN node). Process 1 is the process corresponding to the first AI task, and process 2 is the process corresponding to the second AI task. A process layer exists between process 1, process 2, and the AN protocol layer. The RAN node, in addition to including the AN protocol layer, also has a transport network layer (TNL). The process manager function (PMF) network element includes the transport network layer (TNL) connected to the RAN node and the process layer. The transport network layer (TNL) of the PMF network element also connects to different functional entities (such as…). Figure 4 The diagram shows the connection between functional entity A and functional entity B.

[0122] The Process Management Function (PMF) network element can be integrated into the RAN node, the User Equipment (UE), or configured as an independent network element in the core network. In other words, the PMF network element's functionality can be implemented by the RAN node, the UE, or the core network. Optionally, the PMF network element's functionality can also be implemented by the cloud network.

[0123] In this embodiment, the functional entity is associated with the AI ​​task. For example, the AI ​​task may include, but is not limited to, transceiver, channel estimation, localization, and mobility management. In this application embodiment, the functional entity is used to provide AI services. For example, the AI ​​services may include, but are not limited to, AI model inference services, AI model training services, and AI model training data support services. Furthermore, in this application embodiment, the functional entity may have pre-defined parameters related to the AI ​​task. For example, the parameters related to the AI ​​task may include, but are not limited to, model parameters related to the AI ​​inference task and datasets related to the AI ​​model training task.

[0124] Please refer to Figure 5 The diagram illustrates another communication system that can be used to implement the data transmission method provided in the embodiments of this application, and... Figure 4 Compared to the provided communication system architecture, Figure 5The provided communication system integrates the Process Function (PMF) network element into the User Equipment (UE) or into the RAN node. The system framework no longer needs to set up an additional PMF network element. If the PMF network element is integrated into the RAN node, the RAN node is configured with a process layer.

[0125] It should be noted that the processing layer can be an existing data protocol layer or a newly added data protocol layer. This application embodiment does not limit it. If the processing layer is not a newly added data protocol layer, the processing layer can be set between the application layer and the network access layer.

[0126] The above content describes various wireless communication scenarios involved in this application. It should be understood that the above content is merely an exemplary description of the scenarios in which this application can be applied. This application can also be applied to other application scenarios, which are not limited here.

[0127] The data transmission method and related apparatus provided in the embodiments of this application will be further described below with reference to the accompanying drawings.

[0128] In AI task processing solutions based on wireless access networks, the core network processes the AI ​​task-related data sent by user equipment. Specifically, the Network Data Analysis (NWDAF) network element within the core network executes the AI ​​tasks. However, different AI tasks may require different resources, such as different AI models or different transmission resources. Therefore, a solution based on NWDAF for all AI tasks suffers from insufficient execution efficiency to meet requirements.

[0129] To overcome at least one of the above-mentioned technical problems, embodiments of this application provide a data transmission method for improving the execution efficiency of AI tasks. Figure 6 and Figure 7 A flowchart of a data transmission method is provided. Figure 6 The explanation is based on the first communication device as the executing entity. Figure 7 Indication Figure 4 The communication system architecture shown is for illustrative purposes only.

[0130] It should be noted that the embodiments of this application do not limit the subject executing the data transmission method; the method can be applied to, for example... Figure 6 The provided first communication device can also be applied to, for example Figure 7The provided user equipment (UE), access network RAN ​​equipment, or process processing function (PMF) network element, i.e., the data transmission method provided in this application embodiment, can be applied to both the terminal side and the network side.

[0131] Among them, the terminal side includes, for example, the terminal or the communication / computing module in the terminal, or the circuit or chip in the terminal responsible for communication functions (such as modem chip, also known as baseband chip, or system on chip (SoC) chip containing modem core or system in package (SIP) chip), or the circuit or chip in the terminal responsible for computing and / or communication functions (such as graphics processing unit (GPU), artificial intelligence (AI) processor, or application specific integrated circuit (ASIC)), or the logic node, logic module or software that can realize all or part of the terminal functions.

[0132] Among them, network side, such as network side access network equipment, modules (such as circuits, chips or chip systems) in access network equipment, or logical nodes, logical modules or software that can realize all or part of the functions of access network equipment, or circuits or chips (such as GPUs, AI processors or ASICs) in access network equipment responsible for computing and / or communication functions, or network side core network (CN) equipment.

[0133] Let's combine the following... Figure 6 The schematic diagram illustrating a data transmission method describes the wireless communication process involved in the data transmission method provided in this application embodiment. The method includes the following steps:

[0134] S601: Obtain AI task processing request.

[0135] The AI ​​task processing request indicates the first task identifier and the first input data of the first task. The first task includes a training task and / or an inference task based on a first AI model. The task identifier of the first task is used to uniquely identify the first task. In this embodiment, the form of the task identifier is not limited. The task identifier can be determined based on the process number corresponding to the task, or it can be determined based on the task type or the AI ​​service type corresponding to the first task. The first input data of the first task refers to the data to be processed corresponding to the first task, such as the input data for inference by the AI ​​model, the training input data for the AI ​​model training process, etc. The input data corresponding to the task is different in different AI task situations. In this embodiment, the type of input data is not limited.

[0136] Step S601 above can be understood as the first communication device receiving an AI task processing request from the outside. For example, if the first communication device is an access network device, the access network device receives an AI task processing request sent by the terminal; if the first communication device is a core network element, the core network receives an AI task processing request from the terminal sent by the access network device; or, it can also be understood as the internal implementation of the method execution subject. For example, if the execution subject is the terminal, the physical layer of the terminal receives an AI task processing request sent by the upper layer (such as the application layer). In this embodiment, no limitation is made on the implementation method of obtaining the AI ​​task processing request.

[0137] It should be noted that the embodiments of this application do not limit the way in which the AI ​​task processing request indicates the first task identifier and the first input data of the first task. Specifically, it can be indicated directly, for example, by directly including the first task identifier and the first input data in the AI ​​task processing request; it can also be indicated indirectly, for example, by including other information for indicating the acquisition of the first task identifier and the first input data in the AI ​​task processing request; it can also be indicated by a combination of direct and indirect indication, for example, by directly including the first task identifier and other information for indicating the acquisition of the first input data in the AI ​​task processing request.

[0138] S602: Send first input data to the first functional entity that has deployed the first AI model, and there is a correspondence between the first functional entity and the first task identifier.

[0139] Specifically, there is a correspondence between the first functional entity and the first task identifier. The first functional entity can be determined based on the first task identifier, and the first input data corresponding to the first task is sent to the first functional entity for processing. The first functional entity is equipped with a first AI model, and the first task includes training tasks and / or inference tasks based on the first AI model. Therefore, the first functional entity can meet the processing requirements of the first task.

[0140] Based on the above scheme, there is a correspondence between the task identifier of an AI task and a functional entity. Input data for different AI tasks can be sent to the corresponding functional entity for processing based on the corresponding task identifier. Therefore, this application embodiment realizes the processing of AI tasks based on the correspondence between task identifiers and functional entities. Different AI tasks can be executed by functional entities that provide different AI services, instead of being executed by a single network element (such as the Network Data Analysis Function (NWDAF) network element). This application embodiment improves the execution efficiency of AI tasks through the task identifier and AI task allocation mechanism.

[0141] In this embodiment, the functional entities are associated with AI tasks. For example, AI tasks may include, but are not limited to, transceiver functions, channel estimation, positioning, and mobility management. The functional entities in this embodiment provide AI services. For example, AI services may include, but are not limited to, AI model inference services, AI model training services, and AI model training data support services. Furthermore, the functional entities in this embodiment may have pre-defined parameters related to the AI ​​tasks. For example, AI task-related parameters may include, but are not limited to, model parameters related to AI inference tasks and datasets related to AI model training tasks.

[0142] It should be noted that the functional entities in this application embodiment may be created and deployed to the network by an AI service provider. This application embodiment does not limit the form of the functional entities. The functional entities may be implemented by hardware, software, or a combination of software and hardware.

[0143] In one optional implementation, the data transmission method further includes: determining a first functional entity based on a first task identifier and a first mapping relationship, wherein the first mapping relationship includes the correspondence between the first task identifier and the first functional entity.

[0144] In this embodiment, a first mapping relationship is pre-configured. This first mapping relationship includes the correspondence between task identifiers and functional entities. Different task identifiers can correspond to different functional entities or the same functional entity. Based on the first mapping relationship and the task identifier of the first task in the AI ​​task processing request, the functional entity used to process the first task can be determined. Therefore, this embodiment provides an AI task allocation mechanism, which allows different AI tasks to be allocated to the functional entities corresponding to the task identifiers for execution, thereby improving the execution efficiency of AI tasks.

[0145] In one alternative implementation, the first functional entity is determined by the processing layer based on the first task identifier and the first mapping relationship.

[0146] The processing layer can be an existing data protocol layer or a newly added data protocol layer; this embodiment does not impose any limitations. The processing layer can be located between the application layer and the network access layer. The processing layer is pre-configured with a first mapping relationship, enabling the routing of target functional entities based on task identifiers.

[0147] In one alternative implementation, the data transmission method further includes: acquiring an AI task control request, the AI ​​task control request indicating a first task identifier; and sending the AI ​​task control request to a first functional entity, the AI ​​task control request indicating one of: pausing the first task, restarting the first task, or terminating the first task.

[0148] In this embodiment, an AI task control request can be obtained. This request includes the task's identifier. Based on the AI ​​task control request, the entire lifecycle of the task can be controlled, such as pausing, restarting, and terminating the task. Therefore, by controlling the task, the needs of AI task processing can be responded to promptly, further improving the execution efficiency of AI tasks.

[0149] In this context, task execution refers to the process by which a functional entity processes the input data of a task. This processing can be achieved through processes within the functional entity. Correspondingly, task control can be understood as control of processes within the functional entity, such as process termination, restart, and release. It is understood that in this embodiment, the task identifier can also be replaced with a process identifier.

[0150] Process termination refers to pausing an existing process while preserving or temporarily releasing its associated resources for use by other processes. Once the process restarts, the associated resources are restored. Process restart refers to restarting a paused process and restoring its associated resources. Process release refers to releasing the associated resources and process identifiers of a process for use by other processes.

[0151] It should be noted that the above steps of obtaining the AI ​​task control request can be understood as the first communication device receiving the AI ​​task control request from the outside. For example, if the first communication device is an access network device, the access network device receives the AI ​​task control request sent by the terminal; if the first communication device is a core network element, the core network receives the AI ​​task control request from the terminal sent by the access network device; or, it can also be understood as the internal implementation of the method execution subject. For example, if the execution subject is the terminal, the physical layer of the terminal receives the AI ​​task control request sent by the upper layer (such as the application layer). In this embodiment, the implementation method of obtaining the AI ​​task control request is not limited.

[0152] In one alternative implementation, the data transmission method further includes: within a first time window, a first task identifier is also used to identify a second task, wherein the second task is a different task from the first task.

[0153] The first time window can be determined based on the processing time of the first input data. Within the first time window, the first task identifier can be reused, meaning it can be used to identify other AI tasks, such as the second task. Since AI tasks may require a long processing time during execution, such as high computational complexity or high inference latency, reusing the first task identifier within the processing time of the first input data can improve the utilization rate of task identifier resources.

[0154] Next Figure 4 The communication system architecture shown is for illustrative purposes only. Figure 7 The flowchart illustrating another data transmission method describes the wireless communication process involved in the data transmission method provided in this application embodiment. The method includes the following steps:

[0155] S701: The User Equipment (UE) sends an AI task processing request to the Access Network (RAN) device.

[0156] Specifically, the AI ​​task processing request indicates the first task identifier and the first input data of the first task. The first task includes a training task and / or inference task based on the first AI model. The UE may determine the AI ​​task processing request from an upper layer (such as the application layer). This application embodiment does not limit the implementation method of determining the AI ​​task processing request.

[0157] It should be noted that the relevant explanation of AI task processing requests can be found in the explanation of step 601 of the previous embodiment, and will not be repeated here.

[0158] In one optional implementation, the AI ​​task processing request is determined by the UE's process layer. The process layer can determine the corresponding task identifier for the first task based on the task type of the first task or the AI ​​service type corresponding to the first task. The task identifier is used by the network side to route the corresponding functional entity for the AI ​​task processing request. The process layer adds the determined task identifier of the first task to the header of the data protocol packet of the input data of the first task to obtain the AI ​​task processing request, and sends the AI ​​task processing request to the transport network layer (TNL), so that the TNL sends the AI ​​task processing request to the RAN device.

[0159] In one alternative implementation, the AI ​​task processing request in step S701 above can be carried by a first transmission resource. There is a correspondence between the first transmission resource and the first status identifier of the first task, and the first status identifier indicates the status of the first task.

[0160] Specifically, the UE can also determine a first status identifier based on the status of the first task, and a second mapping relationship is pre-configured in the UE. The second mapping relationship includes the correspondence between the status identifier and the transmission resource. Based on this, the UE can determine the corresponding first transmission resource according to the first status identifier, and send an AI task processing request to the RAN device based on the first transmission resource.

[0161] It should be noted that the configuration method of the second mapping relationship is not limited in the embodiments of this application. For example, the second mapping relationship can be pre-configured to the UE by the PMF network element or other network elements.

[0162] Understandably, the first state identifier is used to indicate the first state of the first task. The first state can be the task execution state of the first task. For example, if the first task is an AI model training task, the first state can include the AI ​​model training stage, the convergence status of the AI ​​model, the data consumption of the AI ​​model, the accuracy of the AI ​​model, etc. Optionally, the first state can also be associated with the task type of the first task, such as: the first task is an AI model training task, or an AI model inference task, etc.

[0163] Because different tasks in different states have different requirements for data transmission—for example, in the early stages of model training, the accuracy requirements for data interaction are not high, while in the later stages of model training, when the model is close to convergence, the accuracy of data transmission will increase accordingly—and because AI recognition tasks for autonomous driving have higher requirements for data transmission latency and reliability compared to image AI creation tasks, the embodiments of this application establish a correspondence between task state identifiers and transmission resources. Based on the task state identifier, the transmission resources used to carry AI task processing requests can be determined. This allows for the determination of different transmission resources based on different task states, with the transmission resources used to carry the data corresponding to the task. This provides suitable transmission resources for AI tasks in different states, thereby improving the execution efficiency of AI tasks.

[0164] It should be noted that the transmission resources corresponding to different status identifiers may be different. For example, the bandwidth parameters, latency parameters, whether retransmission requests are supported, whether high-order modulation is supported, and whether low-order modulation are supported are all possible. In this embodiment, no limitation is made on the transmission resources corresponding to different status identifiers. Those skilled in the art can make selections based on the actual business situation.

[0165] S702: The RAN device forwards the AI ​​task processing request to the PMF network element with process processing function.

[0166] Specifically, after receiving an AI task processing request from the UE, the RAN device forwards the request to the PMF network element, which then dispatches and executes the task. The PMF network element can be integrated into the RAN device, the UE, or configured as an independent network element in the core network. It should be noted that this embodiment uses... Figure 4 The illustrated communication system architecture diagram shows that the PMF network element is set up as an independent network element in the core network. In some alternative implementations, in this embodiment, the PMF network element may not be set up independently; instead, its functions and roles may be integrated into the RAN device or UE, and the data transmission method may be applied to, for example... Figure 5 The communication system architecture shown in this application does not limit the application scenarios of the provided data transmission method.

[0167] S703: The PMF network element determines the first functional entity based on the first mapping relationship and the task identifier of the first task.

[0168] Specifically, the first functional entity is equipped with a first AI model, which is associated with a first task. The first task includes a training task and / or an inference task based on the first AI model.

[0169] The PMF network element is pre-configured with a first mapping relationship, which includes the correspondence between task identifiers and functional entities. Different task identifiers can correspond to different functional entities or the same functional entity. Based on the first mapping relationship and the task identifier of the first task in the AI ​​task processing request, the PMF network element can determine the first functional entity used to process the first task. Therefore, this embodiment of the application provides an AI task allocation mechanism, which allows different AI tasks to be allocated to the functional entities corresponding to the task identifiers for execution, thereby improving the execution efficiency of AI tasks.

[0170] In this embodiment, the configuration method of the first mapping relationship is not limited. It can be determined through negotiation between the AI ​​service provider (i.e., the functional entity provider) and the user, and configured to the UE and PMF through the network.

[0171] It should be noted that the functional entities in this application embodiment may be created and deployed to the network by an AI service provider. This application embodiment does not limit the form of the functional entities. The functional entities may be implemented by hardware, software, or a combination of software and hardware.

[0172] S704: The PMF network element sends the first input data to the first functional entity.

[0173] S705: The PMF network element receives the first output data sent by the first functional entity.

[0174] S706: The PMF network element sends the AI ​​task processing results to the RAN equipment.

[0175] S707: The RAN device sends the AI ​​task processing results to the UE.

[0176] In steps S704 to S707 above, after determining the first functional entity, the PMF network element sends first input data to the first functional entity so that the first functional entity can perform the processing of the first task based on the first input data. Next, after processing the first input data, the first functional entity sends the first output data of the first task to the PMF network element. Then, the PMF network element sends the AI ​​task processing result to the RAN device, which includes the first output data of the first task. Finally, the RAN device forwards the AI ​​task processing result sent by the PMF network element to the corresponding UE to complete the AI ​​task processing of this round.

[0177] In one optional implementation, the data transmission method further includes: the UE sending an AI task control request to a first functional entity, the AI ​​task control request indicating a first task identifier, the AI ​​task control request being used to indicate one of: pausing the first task, restarting the first task, and terminating the first task.

[0178] Specifically, when the UE sends an AI task control request to the first functional entity, it means that the UE sends an AI task control request to the RAN device, the RAN device forwards the AI ​​task control request to the PMF, and the PMF further sends the AI ​​task control request to the corresponding first functional entity.

[0179] Therefore, during the execution of the first task, the UE can perform corresponding control over the first task, specifically by pausing, restarting, or terminating it. This control allows for timely response to the demands of AI task processing, further improving the execution efficiency of AI tasks. It can be understood that controlling the first task essentially involves controlling processes within the functional entity, such as terminating, restarting, and releasing processes.

[0180] In one optional implementation, step S707 specifically includes: the RAN device sending first output data of the first task to the UE based on the first transmission resource; or, sending the first output data to the terminal based on the second transmission resource, wherein there is a correspondence between the second transmission resource and the second status identifier, and the second status identifier is used to indicate the second status of the first task.

[0181] Specifically, the RAN device can send AI task processing results to the UE in two ways.

[0182] In the first method, the RAN device sends the AI ​​task processing result to the UE based on the first transmission resource, which is the resource used by the UE to send the AI ​​task processing request. Thus, the RAN device directly sends the AI ​​task processing result to the terminal according to the resource used by the UE to send the AI ​​task processing request, without having to additionally determine the transmission resource used for the AI ​​task processing result, saving the data transmission overhead on the network side.

[0183] In the second approach, the RAN device can send the AI ​​task processing result to the UE based on the second transmission resource. There is a correspondence between the second transmission resource and the second status identifier, which is used to indicate the second status of the first task. That is, the RAN device needs to determine the second status identifier of the first task and determine the second transmission resource used to carry the AI ​​task processing result based on the second status identifier.

[0184] It is understandable that after the first functional entity processes the first input data to obtain the first output data, the state of the first task may change, for example, the convergence status of the corresponding model may change. Therefore, the RAN device determines the transmission resources based on the task's state identifier, which makes the second transmission resources used to carry the AI ​​task processing results more adaptable to the current state of the first task, providing suitable transmission resources for the AI ​​task and thus improving the execution efficiency of the AI ​​task.

[0185] It should be noted that both the first state and the second state refer to the states of the first task. The first state and the second state refer to the states of the first task at different task lifecycles. The first state is the state of the first task when the terminal sends an AI task processing request to the network side, and the second state is the state of the first task when the network side sends the first output data to the terminal. The first state and the second state may be different or the same; correspondingly, the first state identifier and the second state identifier may be the same or different; and correspondingly, the first transmission resource and the second transmission resource may be the same or different.

[0186] In one alternative implementation, the AI ​​task processing result also includes a second state identifier for the first task.

[0187] Specifically, in step S706 above, the PMF network element sends the AI ​​task processing result to the RAN device. This AI task processing result includes not only the first output data of the first task but also a second status identifier for the first task. That is, when the PMF network element receives the first output data sent by the first functional entity, it can also perceive the latest status of the first task and send the second status identifier along with the first output data to the RAN device. Therefore, the RAN device, upon receiving the AI ​​task processing result, can simultaneously know the latest status of the first task and determine the second transmission resource used to carry the AI ​​task processing result based on the second status identifier of the first task and the second mapping relationship.

[0188] It should be noted that the RAN device is pre-configured with a second mapping relationship, which includes the correspondence between status identifiers and transmission resources. Different status identifiers can correspond to different transmission resources. The RAN device can determine the second transmission resource used to carry the AI ​​task processing result based on the second mapping relationship and the second status identifier of the first task in the AI ​​task processing result. In this embodiment, the configuration method of the second mapping relationship is not limited. For example, the second mapping relationship can be pre-configured to the RAN device by the PMF network element or other network elements.

[0189] Optionally, different state identifiers can be indicated by a set of preset indices. Each index corresponds to a predefined set of task-related state parameters, such as the amount of data consumed during model training and the accuracy of the model.

[0190] Figure 8 and Figure 9 The diagrams illustrate three methods for determining transmission resources, each including process 1, process 2, and process 3. Different processes correspond to different AI tasks. For example, process 1 corresponds to the first AI task. Data from different AI tasks is mapped to different air interface bearers through a processing layer. Figure 8 In this process, data from different AI tasks are mapped to different radio bearers (RBs). Figure 9 Data from different AI tasks is mapped to different quality of service (QoS) streams, and the subsequent mapping is completed by the service data adaptation protocol (SDAP) layer based on the air interface protocol stack.

[0191] In one optional implementation, the data transmission method further includes: the UE obtaining priority information between different processes, and the UE multiplexing data of different AI tasks into a protocol data unit (PDU) based on the priority information between different processes.

[0192] Specifically, the PMF can indicate the priority information of different processes in the UE, thereby facilitating the UE to schedule data for different tasks based on the priority information, as shown in the reference. Figure 10 The diagram shows a data structure of a Protocol Data Unit (PDU) corresponding to an AI task processing request. The PDU includes a header and a service data unit (SDU). The header includes the process ID (PID), status, length, and process data for each process.

[0193] Additionally, the data structure of the Protocol Data Unit (PDU) corresponding to the AI ​​task control request may include the process identifier (PID) and control instruction information corresponding to the process. The above description of the data structure is merely an illustrative example of an optional implementation and does not constitute a limitation on the data structure of the PDU corresponding to the AI ​​task processing request and AI task control request in this embodiment. Those skilled in the art can configure the corresponding data structure according to actual business needs.

[0194] To facilitate the differentiation and identification of AI task processing requests and AI task control requests, an indicator can be added to the header of the protocol data unit corresponding to the AI ​​task processing request and the AI ​​task control request. The corresponding bit symbol indicates whether the protocol data unit belongs to an AI task processing request or an AI task control request.

[0195] In one alternative implementation, the data transmission method further includes: the UE acquiring first delay information, which is used to determine the time to receive the AI ​​task processing result.

[0196] The following is combined with Figure 11 The diagram illustrating a latency information transmission process explains the method corresponding to this step. After receiving an AI task processing request, the PMF can determine feedback information and send it to the UE. The feedback information includes first latency information, which indicates the processing duration of the first input data. Based on this first latency information, the UE can determine the detection window for receiving the AI ​​task processing result. Since AI tasks may require a long processing time during execution, such as high computational complexity or high inference latency, the UE does not need to detect and receive the AI ​​task processing result during the processing time of the first input data, thus saving the overhead of AI task execution.

[0197] In one alternative implementation, the data transmission method further includes: within a first time window, a first task identifier is also used to identify a second task, wherein the second task is a different task from the first task.

[0198] Specifically, the first time window can be determined based on the processing time of the first input data, or based on the first latency information involved in the above method. Within the first time window, the first task identifier can be reused, that is, the first task identifier can be used to identify other AI tasks, such as the second task. Since AI tasks may require a long processing time during execution, such as high computational complexity or high inference latency, reusing the first task identifier within the processing time of the first input data can improve the utilization rate of task identifier resources.

[0199] Figure 12A possible schematic block diagram of the communication device involved in an embodiment of this application is shown. For example... Figure 12 As shown, the communication device 1200 may include modules or units for implementing the methods described in the above embodiments. In one possible design, the communication device 1200 includes a processing unit 1202 and a communication unit 1203. Optionally, the communication device 1200 may further include a storage unit 1201 for storing device program code and / or data.

[0200] The communication device 1200 can be a terminal-side device as described in the above embodiments, such as a terminal or a communication module within a terminal, or a circuit or chip within a terminal responsible for communication functions. The communication device 1200 can also be a network-side device as described in the above embodiments, such as a network-side access network device, a module (e.g., a circuit, chip, or chip system) within the access network device, or a logical node, logical module, or software capable of implementing all or part of the functions of the access network device, or a circuit or chip (e.g., a GPU, AI processor, or ASIC) within the access network device responsible for computing and / or communication functions, or a network-side core network (CN) device.

[0201] Processing unit 1202 is used to: acquire an artificial intelligence (AI) task processing request, wherein the AI ​​task processing request indicates a first task identifier of a first task and first input data of the first task, and the first task includes a training task and / or an inference task based on a first AI model;

[0202] The communication unit 1203 is used to send first input data to a first functional entity on which the first AI model is deployed, wherein there is a correspondence between the first functional entity and the first task identifier.

[0203] In one possible design, the processing unit 1202 is further configured to: determine a first functional entity based on a first task identifier and a first mapping relationship, wherein the first mapping relationship includes the correspondence between the first task identifier and the first functional entity.

[0204] In one possible design, the communication unit 1203 is further configured to: acquire an AI task control request, the AI ​​task control request indicating a first task identifier; and send the AI ​​task control request to a first functional entity, the AI ​​task control request indicating one of: pausing the first task, restarting the first task, or terminating the first task.

[0205] In one possible design, the processing unit 1202 is also used to: identify a second task based on a first task identifier within a first time window, wherein the second task is a different task from the first task.

[0206] In one possible design, the processing unit 1202 is specifically used to: receive an AI task processing request from a terminal, the AI ​​task processing request being carried by a first transmission resource, the first transmission resource having a correspondence with a first status identifier, and the first status identifier indicating the first status of the first task.

[0207] In one possible design, the communication unit 1203 is specifically used to: send first output data of a first task to a terminal based on a first transmission resource; or, send first output data to a terminal based on a second transmission resource, wherein there is a correspondence between the second transmission resource and a second status identifier, and the second status identifier is used to indicate the second status of the first task.

[0208] In one possible design, when the communication device 1200 is a terminal or a communication module within a terminal, the function of the processing unit 1202 can be implemented by one or more processors. Specifically, the processor may include a modem chip, or a system-on-a-chip (SoC) chip or a SIP chip containing a modem core. The function of the communication unit 1203 can be implemented by transceiver circuitry.

[0209] In one possible design, when the communication device 1200 is a circuit or chip in a terminal responsible for communication functions, such as a modem chip or a system-on-a-chip (SoC) or SIP chip containing a modem core, the function of the processing unit 1202 can be implemented by a circuit system in the aforementioned chip that includes one or more processors or processor cores. The function of the communication unit 1203 can be implemented by an interface circuit or data transceiver circuit on the aforementioned chip.

[0210] In one possible design, when the communication device 1200 is a terminal or a computing and / or communication module within a terminal, the functionality of the processing unit 1202 can be implemented by one or more processors. Specifically, the processor may include a GPU, or a system-on-a-chip (SoC) or SIP chip containing a GPU. Alternatively, the processor may include an AI processor, or a SoC or SIP chip containing an AI processor. Or, the processor may include an ASIC, or a SoC or SIP chip containing an ASIC. The functionality of the communication unit 1203 can be implemented by transceiver circuitry.

[0211] In one possible design, when the communication device 1200 is a circuit or chip in a terminal responsible for computing and / or communication functions, such as a GPU or a system-on-a-chip (SoC) or SIP chip containing a GPU, an AI processor or a SoC or SIP chip containing an AI processor, or an ASIC or a SoC or SIP chip containing an ASIC, the function of the processing unit 1202 can be implemented by a circuit system in the aforementioned chip that includes one or more processors or processor cores. The function of the communication unit 1203 can be implemented by interface circuits or data transceiver circuits on the aforementioned chip.

[0212] It is understood that the division of units in the above-described device is merely a logical functional division. One function can correspond to one functional unit, or two or more functions can be integrated into one functional unit. In actual implementation, all or some units can be integrated onto a single physical entity, or distributed across different physical entities. Furthermore, the aforementioned functional units can be implemented in hardware, software, or a combination of both. Whether a function is executed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for specific applications, but such implementations should not be considered beyond the scope of this application.

[0213] In one example, the functional unit in any of the above devices may be one or more integrated circuits configured to implement the above methods, such as: one or more application-specific integrated circuits (ASICs), or one or more central processing units (CPUs), one or more microcontroller units (MCUs), one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs), or a combination of at least two of these integrated circuit forms.

[0214] In one example, storage unit 1201 may include random access memory, flash memory, read-only memory, programmable read-only memory or electrically erasable programmable memory and / or registers, etc.

[0215] It should be noted that the methods and corresponding technical effects performed by the modules of the aforementioned communication device 1200 can be found in the descriptions of the method embodiments shown in the foregoing of this application, and will not be repeated here.

[0216] See Figure 13 This is a schematic diagram of the structure of a terminal 1300 provided in an embodiment of this application. The terminal 1300 can correspond to... Figure 1 , Figure 3 , Figure 4 as well as Figure 5 The terminal shown is used to implement the operations of the terminal in the above embodiments. Figure 13 As shown, the terminal includes: one or more antennas 1310, a radio frequency processing system 1320, and a processor system 1330.

[0217] In the downlink or sidelink direction, the RF processing system 1320 receives RF signals through the antenna 1310 and sends the RF-processed signals to the processor system 1330 for further processing. In the uplink or sidelink direction, the processor system 1330 processes the terminal-side information and sends it to the RF processing system 1320, which then processes the signal and transmits it through the antenna 1310.

[0218] In one example, the RF processing system 1320 serves as the communication interface for external communication of the terminal and may include an RF front end (RFFE) 1321 and an RF transceiver 1322. The RFFE 1321 is primarily used for one or more processing operations, such as shaping, passband selection, or gain adjustment, on the RF signals received by the antenna or those to be transmitted through the antenna. It may include one or more components such as RF switches, duplexers, filters, power amplifiers, antenna tuners, and low-noise amplifiers. The RFFE 1321 can be a circuit system composed of multiple discrete components or integrated into one or more chips. The RF transceiver 1322 processes the RF signals received by the RFFE into baseband / IF signals for further processing by the processor system 1330, and processes the baseband / IF signals provided by the processor system 1330 into RF signals for transmission to the RFFE 1321. The baseband / IF signals transmitted between the RF transceiver 1322 and the processor system 1330 can be digital or analog signals. The radio frequency transceiver 1322 can be implemented by one or more chips, which are commonly referred to as radio frequency chips (RFICs).

[0219] In one example, processor system 1330 may include one or more processors for processing signals and executing one or more communication protocols. Optionally, processor system 1330 may also include memory 1336. In one example, the one or more processors include at least one baseband processor 1331 (also known as a modem processor). Memory 1336 is used to store data and / or computer program instructions. Optionally, processor system 1330 may also include one or more application processors 1332 for implementing processing of the terminal operating system and application layer. Application processor 1332 may include, for example, a GPU, AI processor, or ASIC. Optionally, processor system 1330 may also include one or more of a voice subsystem 1333, a multimedia subsystem 1334, or an interface circuit 1335. The voice subsystem 1333 is used to process voice signals, the multimedia subsystem 1334 is used to handle multimedia-related operations, such as video encoding / decoding, image processing, etc., and the interface circuit 1335 is used to implement communication with other terminal components, such as a display 1340, an input device 1350, memory 1360, etc. The aforementioned components in the processor system 1330 can communicate with each other via a bus or communication interface circuit.

[0220] In one example, the processor system 1330 can be packaged as a single processor chip, such as a SoC chip or a SIP chip. In another example, the processor system 1330 can be a system composed of multiple chips, for example, the baseband processor 1331 can be packaged as a single chip, or packaged with part or all of the circuitry of the radio frequency processing system into a single chip.

[0221] In one example, memory 1336 can be on-chip memory, i.e., located on the processor system 1330 chip. In another example, memory 1360 can be off-chip memory, i.e. located outside the processor system 1330 chip.

[0222] In one example, the baseband processor 1331 may include one or more processor cores 13311 and interface circuitry 13314. The one or more processor cores 13311 are used to process signals and execute one or more communication protocols. Optionally, the baseband processor 1331 may also include a memory 13312 for storing at least a portion of the corresponding computer program instructions and / or data. In one example, the one or more processor cores 13311 execute the computer program instructions stored in the memory 13312 to implement the relevant operations (such as steps S601 to S602, steps S701 to S707) in the above method embodiments. In this disclosure, memory 13312 is used to store corresponding computer program instructions and / or data. This can mean that memory 13312 stores all corresponding computer program instructions and / or data for execution by processor core 13311; or it can mean that memory 13312 stores a portion of corresponding computer program instructions and / or data, including the computer program instructions and / or data currently required to be executed by processor core 13311. Memory 13312 can store different portions of computer program instructions and / or data multiple times for execution by processor core 13311 to implement the relevant operations in the above method embodiments. Interface circuit 13314 serves as a communication interface for communication with other components, such as transmitting signals with radio frequency processing system 1320, communicating with other subsystems and related components of processor system 1330 via bus, such as transmitting data control signals with application processor 1332, and transmitting data or computer program instructions with memory 1336 or memory 1360. Optionally, in order to reduce the load on the processor core, a baseband signal processing circuit 13313 can be set to perform at least some baseband signal processing, including one or more of signal demodulation, modulation, encoding or decoding.

[0223] In one example, the communication device provided in this application may be a terminal 1300, a communication module including a processor system 1330 and a radio frequency system 1320, or a baseband processor 1331.

[0224] The processor, processor system, application processor, baseband processor, processor circuit, or processor core mentioned above can be collectively referred to as a processor. The processor may include one or more of the following: central processing unit (CPU), digital signal processor (DSP), microprocessor unit (MPU), microcontroller unit (MCU), graphics processing unit (GPU), field programmable gate array (FPGA), application-specific integrated circuit (ASIC), artificial intelligence processor (AI processor), or neural processing unit (NPU).

[0225] The aforementioned memory may include one or more of the following storage media: random access memory (RAM), static random access memory (SRAM), dynamic random access memory (DRAM), phase-change memory (PCM), resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), cache, register, read-only memory (ROM), flash memory, erasable programmable read-only memory (EPROM), hard disk, etc. In one example, computer program instructions for executing the above embodiments may be stored on non-volatile memory, such as at least a portion of the aforementioned memory 1360 (e.g., one or more of ROM, flash memory, EPROM, or hard disk). When the terminal is running, the corresponding computer program instructions may be partially or wholly loaded onto a memory with a faster transfer speed than the processor, such as at least a portion of memory 1336 and / or memory 13312 (e.g., one or more of RAM, SRAM, DRAM, PCM, RERAM, MRAM, FRAM, cache, or register), for the processor to execute in order to implement the steps in the above method embodiments.

[0226] In one example, the RF transceiver 1322 and the RF front-end 1321 can also be packaged in a single chip. In another example, the RF transceiver 1322, the RF front-end 1321, and the baseband processor 1331 can also be packaged in a single chip.

[0227] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, etc.) containing computer-usable program code.

[0228] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0229] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0230] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0231] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A data transmission method, characterized in that, The method includes: Obtain an AI task processing request, wherein the AI ​​task processing request indicates a first task identifier of a first task and a first input data of the first task, the first task including a training task and / or an inference task based on a first AI model; The first input data is sent to a first functional entity that has deployed the first AI model, and the first functional entity has a corresponding relationship with the first task identifier.

2. The method according to claim 1, characterized in that, The method further includes: The first functional entity is determined based on the first task identifier and the first mapping relationship, wherein the first mapping relationship includes the correspondence between the first task identifier and the first functional entity.

3. The method according to claim 1 or 2, characterized in that, The method further includes: Obtain an AI task control request, wherein the AI ​​task control request indicates the first task identifier; The AI ​​task control request is sent to the first functional entity, and the AI ​​task control request is used to instruct one of: pausing the first task, restarting the first task, and terminating the first task.

4. The method according to any one of claims 1-3, characterized in that, Within the first time window, the first task identifier is also used to identify a second task, which is a different task from the first task.

5. The method according to any one of claims 1-4, characterized in that, The process of obtaining an AI task processing request includes: receiving the AI ​​task processing request from a terminal, wherein the AI ​​task processing request is carried by a first transmission resource, and there is a correspondence between the first transmission resource and a first status identifier, wherein the first status identifier indicates the first status of the first task.

6. The method according to claim 5, characterized in that, The method further includes: Based on the first transmission resource, the first output data of the first task is sent to the terminal; or... The first output data is sent to the terminal based on the second transmission resource, and there is a correspondence between the second transmission resource and the second status identifier, which is used to indicate the second status of the first task.

7. A data transmission method, characterized in that, The method includes: Send an AI task processing request, the AI ​​task processing request indicating a first task identifier and a first input data of the first task, the first task including a training task and / or an inference task based on a first AI model; Receive first output data of the first task from a first functional entity, wherein there is a correspondence between the first functional entity and the first task identifier.

8. The method according to claim 7, characterized in that, The method further includes: Send an AI task control request to the first functional entity. The AI ​​task control request indicates the first task identifier and is used to indicate one of: pausing the first task, restarting the first task, or terminating the first task.

9. The method according to claim 7 or 8, characterized in that, Within the first time window, the first task identifier is also used to identify a second task, which is a different task from the first task.

10. The method according to any one of claims 7-9, characterized in that, The method further includes: Obtain a first status identifier, which is used to indicate the first status of the first task; Sending the AI ​​task processing request includes: sending the AI ​​task processing request based on a first transmission resource, wherein there is a correspondence between the first transmission resource and the first status identifier.

11. The method according to claim 10, characterized in that, The first output data is carried by the first transmission resource; or, the first output data is carried by the second transmission resource, and there is a correspondence between the second transmission resource and the second status identifier, the second status identifier being used to indicate the second status of the first task.

12. A communication device, characterized in that, Includes a unit for performing the method as described in any one of claims 1 to 6.

13. A communication device, characterized in that, Includes units for performing the method as described in any one of claims 7 to 11.

14. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program or instructions that, when executed, cause the method as claimed in any one of claims 1 to 6, or claims 7 to 11, to be implemented.

15. A computer program product, characterized in that, Includes a computer program or instructions that, when executed, cause the method as claimed in any one of claims 1 to 6, or claims 7 to 11, to be implemented.

Images