Model function alignment method in communication system and communication apparatus

Abstract

A model function alignment method in a communication system and a communication apparatus. The method comprises: a terminal acquires a first function supported by a network device; the terminal sends first information to the network device, wherein the first information indicates a second function and / or a third function, the second function and the third function are associated with a first task, and the first function comprises the second function and does not comprise the third function; and the terminal acquires a computation result of the first task, the computation result being determined on the basis of AI models of the second function and the third function. In the method, distributed reasoning of functions can be realized, and functions limited by a capability of a terminal side are implemented on a network device side, thereby improving the user experience.

Description

Methods and communication devices for model function alignment in communication systems

[0001] This application claims priority to Chinese Patent Application No. 202410977293.2, filed on July 19, 2024, entitled “Method and Communication Apparatus for Model Function Alignment in a Communication System”, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of communications, and more specifically, to a method and apparatus for model-function alignment in a communication system. Background Technology

[0003] Currently, the standard discusses the deployment of artificial intelligence (AI) models on the terminal side. When AI models are deployed on the terminal side, the advantage is that the terminal side is close to the data source, which can reduce data transmission latency and bandwidth consumption. However, due to limitations in terminal side computing power, storage, and other factors, the functions corresponding to the AI ​​model cannot be implemented on the terminal side or the implementation effect is poor, which in turn affects the user experience. Summary of the Invention

[0004] This application provides a communication method and a communication device that can realize distributed reasoning of functions and improve user experience.

[0005] In a first aspect, embodiments of this application provide a communication method that can be applied to a terminal side, such as a terminal or a communication / processing 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 processing functions (such as a graphics processing unit (GPU)). Taking the application of this method to a terminal as an example, in this method, the terminal obtains a first function supported by a network device; the terminal sends first information to the network device, the first information indicating a second function and / or a third function, the second function and the third function being associated with a first task, the first function including the second function but not including the third function; the terminal obtains the calculation result of the first task, the calculation result of the first task being determined based on an AI model of the second function and the third function.

[0006] The step of the terminal obtaining the first function supported by the network device can also be described as: the terminal obtains information #1, which indicates that the network device supports the first function. In other words, this step describes how the terminal determines which functions the network device currently supports.

[0007] The second and third functions are associated with the first task. This can be understood as the first task being decomposed, and the resulting functions include both the second and third functions. Furthermore, the first function includes the second function but excludes the third function; this can be understood as the second function being a function supported by the network device, and the third function being a function not supported by the network device. In one possible scenario, if the network device supports all functions associated with the first task, then the third function does not exist. In this case, the calculation result of the first task is determined based on the AI ​​model of the second function. In another possible scenario, if the network device does not support all functions associated with the first task, then the second function does not exist. In this case, the calculation result of the first task is determined based on the AI ​​model of the third function.

[0008] In the above technical solution, the terminal can align with the network device to complete the first task based on the functions supported by the acquired network device. This allows for distributed reasoning of the first task (i.e., different devices implementing different functions of the first task) between the terminal and the network device. Based on distributed reasoning, functions limited by the terminal's capabilities can be implemented on the network device side, thereby improving the user experience.

[0009] In one possible design, the first function is defined by functional information, and the first function corresponds to a functional identifier. The functional information indicates one or more of the following of the first function: the dimension of the input information, the modality of the input information, the description of the input information, the dimension of the output information, the modality of the output information, the description of the output information, accuracy, computation latency, memory usage, computing power requirements, and functional description.

[0010] In one possible design, the first information is also used to indicate the execution order of the second and third functions.

[0011] It is understandable that after aligning the network device and terminal model functions, it is also necessary to divide the multiple functions corresponding to the first task into functions, that is, to determine which functions of the multiple functions are executed by the terminal, network device, and server respectively. Therefore, the terminal can also inform the network device of the execution order of the multiple functions, so that each device can execute its corresponding function in sequence based on this order during distributed inference.

[0012] In one possible design, the method further includes: the terminal sending second information to the network device, the second information indicating the functions supported by the terminal, the functions supported by the terminal being included in the second function and the third function.

[0013] In the above technical solution, the terminal reports the functions it supports among the functions associated with the first task to the network device. The network device can jointly determine the method of distributed inference of the first task (i.e., the function division method of the first task) based on which functions it supports in the first task and which functions the terminal supports in the first task.

[0014] In one possible design, the method further includes: the terminal receiving first configuration information from a network device, the first configuration information indicating the available functions of the terminal, the available functions being included in a second function and a third function; the terminal obtaining the calculation result of a first task, including: the terminal obtaining the calculation result of the first task based on the AI ​​model corresponding to the available functions.

[0015] In the above technical solution, the network device can inform the terminal in advance of the functions that may be implemented on the terminal side in the first task corresponding to the function. The terminal only needs to reserve enough resources for the AI ​​model corresponding to the function indicated by the first configuration information, thereby saving terminal resources.

[0016] In one possible design, the terminal obtains a first function supported by the network device, including: the terminal receiving third information from the network device; and the terminal obtaining the first function based on the third information.

[0017] In one possible design, the method further includes: the terminal sending a first request message to the network device, the first request message being used to request a first function.

[0018] Secondly, embodiments of this application provide a communication method that can be applied to a network device, such as a network device, a module (e.g., a circuit, chip, or chip system) within the network device, or a logical node, logical module, or software capable of implementing all or part of the functions of the network device. Taking the application of this method to a network device as an example, in this method, the network device determines that it has received first information from a first device. The first information indicates a second function and / or a third function, the second and third functions are associated with a first task, and the first function includes the second function but does not include the third function. The first function is a function supported by the network device, and the first device is a terminal or a server. The network device obtains a second calculation result, which is determined based on an AI model corresponding to the second function.

[0019] For the beneficial effects of the second aspect, please refer to the description of the first aspect, which will not be repeated here.

[0020] In one possible design, the first function is defined by functional information, and the first function corresponds to a functional identifier. The functional information indicates one or more of the following of the first function: the dimension of the input information, the modality of the input information, the description of the input information, the dimension of the output information, the modality of the output information, the description of the output information, accuracy, computation latency, memory usage, computing power requirements, and functional description.

[0021] In one possible design, the first information is also used to indicate the execution order of the second and third functions.

[0022] In one possible design, the method further includes: the network device receiving second information from the terminal, the second information indicating the functions supported by the terminal, the functions supported by the terminal being included in the second function and the third function.

[0023] In one possible design, the method further includes: the network device sending first configuration information to the terminal, the first configuration information indicating the available functions of the terminal, the available functions being included in the second and third functions.

[0024] In one possible design, the method further includes: the network device sending third information to the first device, the third information indicating the first function.

[0025] In one possible design, the method further includes: the network device receiving a first request message from the first device, the first request message being used to request a first function.

[0026] Thirdly, embodiments of this application provide a communication method that can be applied to a server side, such as a server or a module within a server (e.g., a circuit, chip, or chip system), or a logical node, logical module, or software capable of implementing all or part of the server's functions. Taking the application of this method to a server as an example, in this method, the server obtains a first function supported by a network device; the server sends first information to the network device, the first information indicating a second function and / or a third function, the second and third functions being associated with a first task, the first function including the second function but not including the third function; the server obtains a first calculation result, the first calculation result being determined based on the AI ​​model corresponding to the second and / or third functions.

[0027] For the beneficial effects of the third aspect, please refer to the description of the first aspect, which will not be repeated here.

[0028] In one possible design, the first function is defined by functional information, and the first function corresponds to a functional identifier. The functional information indicates one or more of the following of the first function: the dimension of the input information, the modality of the input information, the description of the input information, the dimension of the output information, the modality of the output information, the description of the output information, accuracy, computation latency, memory usage, computing power requirements, and functional description.

[0029] In one possible design, the first information is also used to indicate the execution order of the second and third functions.

[0030] In one possible design, the server obtains a first function supported by the network device, including: the server receiving third information from the network device; and the server obtaining the first function based on the third information.

[0031] In one possible design, the method further includes: the server sending a first request message to the network device, the first request message being used to request a first function.

[0032] Fourthly, 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.

[0033] Fifthly, 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.

[0034] Sixthly, this application provides a communication device that has the functions of the third aspect above. For example, the communication device includes modules, units, or means that perform the operations involved in the third aspect above. These modules, units, or means can be implemented by software, hardware, or a combination of software and hardware.

[0035] In a seventh aspect, 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. 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. 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.

[0036] In one possible design, the processor is used to communicate with other devices or components through the interface circuit.

[0037] In one possible design, the communication device may also include the memory.

[0038] The aforementioned communication device may be a terminal, or a communication / processing module in the terminal, or a chip in the 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 the terminal responsible for processing functions (such as a GPU).

[0039] Eighthly, 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 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 second 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.

[0040] In one possible design, the processor is used to communicate with other devices or components through the interface circuit.

[0041] In one possible design, the communication device may also include the memory.

[0042] The aforementioned communication device may be a network device, or a module (such as a circuit, chip, or chip system) in a network device, or a logical node, logical module, or software that can realize all or part of the functions of a network device.

[0043] Ninthly, 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 computer program or instructions necessary to implement the functions described in the third 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 third 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.

[0044] In one possible design, the processor is used to communicate with other devices or components through the interface circuit.

[0045] In one possible design, the communication device may also include the memory.

[0046] The aforementioned communication device may be a server, or a module within a server (such as a circuit, chip, or chip system), or a logical node, logical module, or software capable of implementing all or part of the server's functions.

[0047] In a tenth aspect, a computer-readable storage medium is provided that stores program code for execution by a device, the program code including methods for performing any of the foregoing aspects or their implementations.

[0048] In an eleventh aspect, a computer program product containing instructions is provided, which, when run on a computer, causes the computer to perform the method provided by any of the above aspects or their implementations.

[0049] In a twelfth aspect, a communication system is provided, including at least one of the terminal, network device, and server described above. Attached Figure Description

[0050] Figure 1 is a schematic diagram of a communication system applicable to an embodiment of this application.

[0051] Figures 2 and 3 are schematic diagrams of possible application frameworks in a communication system.

[0052] Figures 4 and 5 are schematic flowcharts of the communication method provided in this application.

[0053] Figure 6 is a schematic block diagram of a communication device provided in an embodiment of this application.

[0054] Figure 7 is a schematic diagram of the structure of a terminal provided in an embodiment of this application. Detailed Implementation

[0055] The technical solutions in the embodiments of this application will now be described with reference to the accompanying drawings.

[0056] To facilitate understanding of the above embodiments provided in this application, the following points are made:

[0057] 1) In this application, unless otherwise specified or in case of logical conflict, the terms and / or descriptions of different embodiments are consistent and can be referenced by each other. The technical features of different embodiments can be combined to form new embodiments according to their inherent logical relationship.

[0058] 2) In this application, the terms "system" and "network" are used interchangeably. "At least one" means one or more, and "more than one" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. In the textual description of this application, the character " / " generally indicates that the preceding and following related objects 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 can mean: a, or, b, or, c, or, a and b, or, a and c, or, b and c, or, a, b, and c. Here, a, b, and c can be single or multiple.

[0059] 3) In the embodiments of this application, ordinal numbers such as "first" and "second" are used to distinguish multiple objects and are not used to limit the size, content, order, timing, priority, or importance of the multiple objects. For example, the first instruction information and the second instruction information can be the same information or different information, and such names do not indicate differences in the content, size, application scenario, sending / receiving end, priority, or importance of the two messages. In addition, the numbering of steps in the various embodiments described in this application is only to distinguish different steps and is not used to limit the order of steps.

[0060] 4) In this application, descriptions such as “when…”, “under the circumstances of…” and “if” all refer to the fact that the device will make corresponding processing under certain objective circumstances. They are not time limits, nor do they require the device to make a judgment action when it is implemented, nor do they mean that there are other limitations.

[0061] 5) In this application, "instruction" or "for instruction" can include both direct and indirect instruction. When describing an instruction as being used to instruct A, it may include whether the instruction directly instructs A or indirectly instructs A, but does not necessarily mean that the instruction carries A.

[0062] The indication methods involved in the embodiments of this application should be understood to cover various methods that enable the party to be indicated to obtain the information to be indicated. The information to be indicated can be sent as a whole or divided into multiple sub-information and sent separately. Moreover, the sending period and / or sending time of these sub-information can be the same or different. This application does not limit the sending method, for example.

[0063] The "instruction information" in the embodiments of this application can be an explicit instruction, that is, a direct instruction through signaling, or an instruction obtained by combining other rules or parameters with the parameters indicated by the signaling, or by deduction. It can also be an implicit instruction, that is, an instruction obtained based on rules or relationships, or based on other parameters, or by deduction. This application does not specifically limit it in this regard.

[0064] 6) The “protocol” used in this application may refer to standard protocols in the field of communications, such as fourth-generation (4G) protocols. th Generation 4G network, fifth generation (5G) network th This application does not limit the scope to network protocols such as generation (5G), NR, 5.5G, and related protocols applied in future communication networks.

[0065] 7) In this application, "communication" can also be described as "data transmission", "information transmission", "data processing", etc. "Transmission" includes "sending" and "receiving".

[0066] 8) In this application, "sending information" can be understood as one device sending information to another device, or it can also be understood as one logical module within a device sending information to another logical module. For example, "network device sending information" can be understood as a network device sending information to another device (such as a terminal or server), or it can be understood as logical module 1 in the network device sending information to logical module 2 in the network device.

[0067] In this application, "receiving information" can be understood as one device receiving information from another device, or it can also be understood as a logical module within a device receiving information from another logical module. For example, "network device receiving information" can be understood as a network device receiving information from another device (such as a terminal), or it can be understood as logical module 1 in the network device receiving information from logical module 2 in the network device.

[0068] In this application, phrases such as "sending information to... (e.g., a terminal)" or related illustrations in the accompanying drawings can be understood as indicating that the destination of the information is a terminal. This can include sending information directly or indirectly to a terminal. Similarly, phrases such as "receiving information from... (e.g., a terminal)," "receiving information from... (e.g., a terminal)," or "receiving information sent by (e.g., a terminal)," or related illustrations in the accompanying drawings, can be understood as indicating that the source of the information is a terminal. This can include receiving information directly or indirectly from a terminal. Information may undergo necessary processing between the source and destination, such as format changes, but the destination can understand the valid information from the source. Similar expressions in this application can be interpreted similarly and will not be elaborated further here.

[0069] 9) The terms “comprising” and “having”, and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus.

[0070] The technical solutions of this application embodiment can be applied to various communication systems, such as: Long Term Evolution (LTE) systems, LTE Frequency Division Duplex (FDD) systems, LTE Time Division Duplex (TDD) systems, 5th Generation (5G) systems, or New Radio (NR) and future communication systems, vehicle-to-other devices (V2X), where V2X can include vehicle-to-network (V2N), vehicle-to-vehicle (V2V), vehicle-to-infrastructure (V2I), vehicle-to-pedestrian (V2P), etc., long term evolution-vehicle (LTE-V) technology for vehicle-to-everything (V2V), vehicle-to-everything (V2X), machine-type communication (MTC), Internet of Things (IoT), long term evolution-machine (LTE-M) technology for machine-to-machine (M2M), etc.

[0071] Figure 1 is a schematic diagram of a communication system applicable to an embodiment of this application. It is understood that Figure 1 is a possible, non-limiting system schematic diagram. As shown in Figure 1, the communication system 10 includes a radio access network (RAN) 100 and a core network (CN) 200. RAN 100 includes at least one RAN node (110a and 110b in Figure 1, collectively referred to as 110) and at least one terminal (120a-120j in Figure 1, collectively referred to as 120). RAN 100 may also include other RAN nodes, such as wireless relay devices and / or wireless backhaul devices (not shown in Figure 1). Terminal 120 is wirelessly connected to RAN node 110. RAN node 110 is wirelessly or wired connected to core network 200. The core network equipment in core network 200 and RAN node 110 in RAN 100 can be different physical devices, or they can be the same physical device integrating core network logical functions and radio access network logical functions.

[0072] 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 radio access network (CRAN), or a wireless fidelity (WiFi) system. RAN 100 can also be a communication system that integrates two or more of the above systems.

[0073] RAN node 110, sometimes also referred to as access network equipment, RAN entity, access node, or network equipment, 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. Unless otherwise specified in this application, RAN nodes and network equipment can be interchanged. In some scenarios, the roles of RAN node 110 and terminal 120 are relative. For example, network element 120i in Figure 1 can be a helicopter or drone, which can be configured as a mobile base station. For terminals 120j accessing RAN 100 through network element 120i, network element 120i is a base station; but for base station 110a, network element 120i is a terminal. RAN node 110 and terminal 120 are sometimes both referred to as communication devices. For example, network elements 110a and 110b in Figure 1 can be understood as communication devices with base station functions, and network elements 120a-120j can be understood as communication devices with terminal functions.

[0074] In one possible scenario, the RAN node can be a base station, an evolved NodeB (eNodeB), an access point (AP), a transmission reception point (TRP), a next-generation NodeB (gNB), a base station in a future mobile communication system, or an access node in a WiFi system. The RAN node can be a macro base station (as shown in Figure 1, 110a), a micro base station or indoor station (as shown in Figure 1, 110b), a relay node or donor node, or a radio controller in a CRAN scenario. Optionally, 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, as well as 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 RAN node's functions.

[0075] In another possible scenario, multiple RAN nodes collaborate to assist the terminal in achieving wireless access, with each RAN node performing 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 separate entities 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).

[0076] In different systems, CU (or CU-CP and CU-UP), DU, or RU may have different names, but those skilled in the art will understand their meaning. For example, in an ORAN system, CU can also be called O-CU (open CU), DU can also be called O-DU, CU-CP can also be called O-CU-CP, CU-UP can also be called O-CU-UP, and RU can also be called O-RU. For ease of description, this application uses CU, CU-CP, CU-UP, DU, and RU as examples. 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 and hardware modules.

[0077] A terminal can be a device or module that accesses the aforementioned communication system and has corresponding communication functions. A terminal can also be called a user equipment (UE), mobile station, mobile terminal, etc. Terminals can be widely used in various scenarios, such as device-to-device (D2D), vehicle-to-everything (V2X) communication, machine-type communication (MTC), Internet of Things (IoT), virtual reality, augmented reality, industrial control, autonomous driving, telemedicine, smart grids, smart furniture, smart offices, smart wearables, smart transportation, smart cities, etc. Terminals can be mobile phones, tablets, computers with wireless transceiver capabilities, wearable devices, vehicles, drones, helicopters, airplanes, ships, robots, robotic arms, smart home devices, transportation vehicles with wireless communication capabilities, communication modules, etc. The embodiments of this application do not limit the device form of the terminal. A terminal typically contains a communication module, circuit, or chip that performs the corresponding communication functions. The terminal can also be configured with program instructions for performing the corresponding communication functions.

[0078] To support AI technology in wireless networks, AI nodes may also be introduced into the network.

[0079] AI nodes can be deployed in one or more of the following locations within the communication system: RAN nodes, terminals, or core network elements, etc. Alternatively, AI nodes can be deployed independently, for example, in a location other than any of the above-mentioned devices, such as in the host or cloud server of an over-the-top (OTT) system. AI nodes can communicate with other devices in the communication system, which can be one or more of the following: RAN nodes, terminals, or core network elements, etc.

[0080] 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.

[0081] 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.

[0082] AI nodes can be AI network elements or AI modules.

[0083] Figure 2 illustrates a possible application framework in a communication system. As shown in Figure 2, network elements in the 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 (only one is shown in Figure 2 for clarity). An access network node can be a single RAN node or can include multiple RAN nodes, for example, including CU and DU. The CU and / or DU can also be equipped with one or more AI modules. A CU can also be split into CU-CP and CU-UP, with one or more AI modules configured in the CU-CP and / or CU-UP.

[0084] 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.

[0085] In one example, the neural network mentioned above can be a deep neural network (DNN), a convolutional neural network (CNN), a recurrent neural network (RNN), or a generative adversarial network (GAN).

[0086] 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.

[0087] 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.

[0088] 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.

[0089] 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.

[0090] 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.

[0091] Figure 3 illustrates another possible application framework in a communication system. As shown in Figure 3, the communication system includes a RAN intelligent controller (RIC). For example, the RIC can be the AI ​​modules 117 and 118 shown in Figure 2, 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.

[0092] 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, 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.

[0093] 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, 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.

[0094] 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), while non-real-time RICs can be set in OAM, cloud servers, core network devices, or other network devices.

[0095] To facilitate understanding of the embodiments of this application, some basic concepts involved in this application will be briefly explained.

[0096] 1. AI: This refers to enabling machines to learn, accumulate experience, and solve problems that humans can solve through experience, such as natural language understanding, image recognition, and chess. Artificial intelligence can be understood as the intelligence exhibited by machines created by humans. Generally, AI refers to the technology of using computer programs to represent human intelligence. The goals of AI include understanding intelligence by constructing computer programs that demonstrate symbolic reasoning or logical reasoning.

[0097] 2. Machine Learning: This is one implementation of artificial intelligence. Machine learning is a method that endows machines with the ability to perform functions that cannot be done directly through programming. In practical terms, machine learning is a method that uses data to train a model and then uses the model to make predictions. There are many machine learning methods, such as neural networks (NNs), decision trees, and support vector machines. Machine learning theory mainly involves designing and analyzing algorithms that allow computers to learn automatically. Machine learning algorithms are a class of algorithms that automatically analyze data to obtain patterns and use these patterns to predict unknown data.

[0098] 3. AI Model: An AI model is an algorithm or computer program that enables AI functionality. It represents the mapping relationship between the model's input and output; in other words, it's a function model that maps a certain dimension of input to a certain dimension of output. The parameters of this function model can be obtained through machine learning training. For example, f(x) = ax 2+b is a quadratic function model, which can be viewed as an AI model. a and b are the parameters of this AI model, and a and b can be obtained through machine learning training.

[0099] It is understood that AI models can be implemented as hardware circuits, software, or a combination of both; there are no restrictions. Non-restrictive examples of software include: program code, program, subroutine, instructions, instruction sets, code, code segments, software modules, applications, or software applications, etc. AI models can be neural networks, linear regression models, decision tree models, support vector machines (SVM), Bayesian networks, Q-learning models, or other machine learning models.

[0100] 4. Dataset: The data used for model training, model validation, or model testing in machine learning. The quantity and quality of the data will affect the effectiveness of machine learning. The dataset can include training datasets and inference data.

[0101] The technical problem to be solved and the technical solution adopted in this application are described below.

[0102] AI node-based inference can be categorized into three types from an intelligent deployment perspective: edge computing, device computing, and cloud computing. Simply put, edge computing is computation performed on devices or terminals, edge computing is computation performed on edge nodes closer to the user's device, and cloud computing is a model that provides computing and storage resources via the internet. Currently, standards discuss the deployment of AI models on the device side. The advantage of deploying AI models on the device side is that the device is close to the data source, which can reduce data transmission latency and bandwidth consumption. However, limitations in edge computing power and storage may result in a limited number of AI model parameters that the device can process. For example, this could lead to inability to perform computations, low accuracy of computation results, or high computation latency, thus preventing the corresponding functions of the AI ​​model from being implemented on the device side or achieving poor results, thereby affecting the user experience.

[0103] In view of this, this application proposes a communication method that can effectively solve the above-mentioned technical problems. An embodiment of the method proposed in this application will be described below with reference to FIG4.

[0104] It is understood that this application uses terminals, network devices, and server / core network devices as examples to illustrate the execution of the interaction, but this application does not limit the execution subject of the interaction. For example, the method executed by the network device in this application can also be implemented by a module (e.g., circuit, chip, or chip system) in the network device, or a logical node, logical module, or software that can implement all or part of the functions of the network device; the method executed by the terminal in this application can also be implemented by a communication / processing module in the terminal or a circuit or chip (e.g., a modem chip (also known as a baseband chip), or a SoC chip containing a modem core, or a SIP chip, or a GPU) in the terminal responsible for communication / processing functions; the method executed by the server (or core network device) in this application can also be implemented by a module (e.g., circuit, chip, or chip system) in the server (or core network device), or a logical node, logical module, or software that can implement all or part of the functions of the server (or core network device).

[0105] Figure 4 is a schematic flowchart of a communication method 400 provided in this application. The method includes the following steps.

[0106] S410, the first device acquires the first function supported by the network device.

[0107] For example, the first device is a terminal, a server, or a core network device. In this application, the operation of the server and the core network device is the same; for ease of description, the following description uses a server as an example.

[0108] The step of the terminal obtaining the first function supported by the network device can also be described as: the terminal obtains information #1, which indicates that the network device supports the first function. That is, this step describes the terminal determining which functions the network device currently supports. The network device supporting the first function can be understood as the network device storing or deploying an AI model capable of implementing the first function. The number of AI models corresponding to the first function can be one or more; that is, the network device stores or deploys one or more AI models capable of implementing the first function. It can be understood that all one or more AI models can implement the first function, but the capabilities of these AI models in implementing the first function may differ. For example, all one or more AI models can implement image recognition, but the image recognition accuracy of these AI models may differ.

[0109] It can be understood that the first function can be viewed as a function supported by the network device. Therefore, S410 can also be described as: the terminal obtains a first set of functions supported by the network device, the first set of functions indicating one or more functions supported by the network device. The first set of functions includes the first function.

[0110] In this application, an AI model can perform one function, or an AI model can perform multiple different functions; this application does not impose any restrictions on this.

[0111] In this application, a function can be implemented by one AI model, or a function can be implemented by multiple AI models together, or multiple AI models can implement the same function independently. This application does not impose any restrictions on this.

[0112] For ease of description, the following explanation will use the first function as an example to illustrate the functions supported by the network device.

[0113] In this application, the first function can be defined by function information, which indicates one or more of the following parameters of the first function: function description parameters, input information parameters, output information parameters, performance parameters, and function identifier. The above parameters are described in detail below.

[0114] (1) The function description parameter of the first function is used to describe the specific functions that the first function can achieve.

[0115] For example, the first function could be image recognition, or speech understanding, or text-to-image generation.

[0116] In one possible implementation, the function description parameter uses language to describe the specific function. For example, if the first function is image recognition, the corresponding function description parameter would be the language description "identify objects of interest in an image".

[0117] Another possible implementation method is to standardize or predefine the definition relationship between functions and function identifiers (IDs). For example, the function ID for image recognition can be predefined as 0001, the function ID for speech understanding as 0002, and the function ID for text-to-image generation as 0003. This mapping relationship is aligned across various computing nodes. For example, if the first function is image recognition, then the corresponding function description parameter is "0001".

[0118] (2) The input information parameters of the first function are used to describe the input information supported by the AI ​​model corresponding to the first function.

[0119] For example, the input information parameters may include one or more of the following: the dimension of the input information, the modality of the input information, and the description of the input information.

[0120] Let's take image recognition as an example. For instance, the AI ​​model for the first function expects the input modality to be an image. If the input is text or other modalities, the AI ​​model for the first function cannot process it. The AI ​​model for the first function can support processing images of 360P, 480P, 720P, and 1080P. 1080P corresponds to a 1080*768 matrix. Therefore, the input modality for the AI ​​model for the first function is 360P, 480P, 720P, 1080P, or 360P~1080P. The input information is described as a static image.

[0121] (3) The output information parameter of the first function is used to describe the output information supported by the AI ​​model corresponding to the first function.

[0122] For example, the output information parameters may include one or more of the following: the dimension of the output information, the modality of the output information, and the description of the output information.

[0123] Let's continue with the example of image recognition as the first function. For instance, the AI ​​model corresponding to the first function expects an image as input and labels the object of interest. If other output modalities are required, the AI ​​model corresponding to the first function cannot handle them. Furthermore, the AI ​​model corresponding to the first function can support processed images in dimensions of 360P, 480P, 720P, and 1080P. Therefore, the output dimensions of the AI ​​model corresponding to the first function will be 360P, 480P, 720P, 1080P, or 360P~1080P. The output information can be described as a static image plus the recognition result.

[0124] (4) The performance parameters of the first function are used to describe the performance of the AI ​​model corresponding to the first function.

[0125] For example, performance parameters include one or more of the following: accuracy, computation latency, video memory usage, and computing power requirements.

[0126] It is understandable that the first function may correspond to multiple AI models. Therefore, the performance parameters of the first function can be used to describe the performance range of all AI models that implement the first function. This range can be understood as the interval between the maximum and minimum performance values, or as the minimum or maximum performance value. The following example illustrates this using four AI models, A, B, C, and D, that can implement the first function on the network device side.

[0127] In one possible implementation, the four AI models A, B, C, and D corresponding to the first function correspond to a performance range for the same performance.

[0128] For example, given the same standard dataset, if A has an accuracy of 90%, B has an accuracy of 74%, C has an accuracy of 50%, and D has an accuracy of 99%, then the accuracy of the AI ​​model corresponding to the first function can be described as 50% to 99%, or it can be described as 99% (maximum).

[0129] For example, given the computation latency range, A has a computation latency of 10ms; B has a computation latency of 20ms; C has a computation latency of 40ms; and D has a computation latency of 100ms. Therefore, for this computation node, the computation latency of the AI ​​model corresponding to the first function can be described as 10ms to 100ms, or it can be described as 10ms (the minimum value).

[0130] In another possible implementation, the four AI models A, B, C, and D corresponding to the first function each correspond to a performance value for the same performance.

[0131] For example, Model A has an accuracy of 90%, uses 1GB of video memory, requires 1 TFLOPS of computing power, and has a computation latency of 10ms; Model B has an accuracy of 74%, uses 2GB of video memory, requires 2 TFLOPS of computing power, and has a computation latency of 20ms; Model C has an accuracy of 50%, uses 4GB of video memory, requires 4 TFLOPS of computing power, and has a computation latency of 40ms; and Model D has an accuracy of 99%, uses 3GB of video memory, requires 3 TFLOPS of computing power, and has a computation latency of 100ms. Therefore, the accuracy, computation latency, video memory usage, and computing power requirement corresponding to the first function are {(90%, 1GB, 1TFLOPS, 10ms), (74%, 2GB, 2TFLOPS, 20ms), ..., ...}.

[0132] (5) The function identifier of the first function is used to uniquely indicate the first function. In this application, different functions supported by the network device correspond to different function identifiers.

[0133] For example, the first device can obtain the first function supported by the network device in the following two ways.

[0134] In one possible implementation, the information regarding the functions supported by the network device can be predefined. For example, this information is stored in a first device or another device readable by the first device, and the first device can determine the functions supported by the network device based on the stored information. For example, Table 1 can be predefined, including function information for functions #1, #2, and #3 supported by the network device.

[0135] Table 1

[0136] In another possible implementation, the first device receives third information from the network device and determines the functions supported by the network device based on the third information. For example, the third information includes Table 1, whereby the network device sends a list of its supported functions to the first device in a tabular format. For example, if the first device is a terminal, the third information can be carried in a radio resource control (RRC) message or an application (APP) message. For example, if the first device is a server, the third information can be carried in a non-access stratum (NAS) message.

[0137] Optionally, in this implementation, before the network device sends the third information to the first device, the first device may send a first request message to the network device, the first request message being used to request the network device to inform the network device of the functions it supports; or, before the network device sends the third information to the first device, the network device may determine, based on its own implementation, whether to inform the first device of the functions it supports.

[0138] S420: The first device sends first information to the network device. Correspondingly, the network device receives the first information from the first device.

[0139] The first information indicates the second function and / or the third function, the second function and the third function are associated with the first task, and the first function includes the second function but does not include the third function.

[0140] Optionally, before sending the first information, the first device can functionally decompose the first task to be completed. For example, if the first task is to instruct an intelligent robot (i.e., a terminal) to pick up a package, this task requires understanding human instructions and the visual environment. Therefore, the task can be decomposed into two functions: understanding the visual environment and understanding human instructions (i.e., understanding human language). The second and third functions are associated with the first task; this can be understood as the functions obtained after decomposing the first task including the second and third functions. Furthermore, the first function includes the second function but does not include the third function; this can be understood as the second function being a function supported by the network device, and the third function being a function not supported by the network device.

[0141] It can be understood that the second function can be considered as a function supported by the network device, and the third function can be considered as a function not supported by the network device. Therefore, S420 can also be described as follows: the first device sends first information to the network device, the first information indicating a second function set and / or a third function set, the second function set and the third function set being associated with the first task, all functions in the second function set being functions supported by the network device, and all functions in the third function set being functions not supported by the network device. Specifically, the functions in both the second and third function sets are associated with the first task, the second function set includes the second function, and the third function set includes the third function.

[0142] For example, if the first device is a terminal, the first information can be carried in an RRC message or an APP message. For example, if the first device is a server, the first information can be carried in a NAS message.

[0143] The first piece of information will be explained in the following sections, with different scenarios.

[0144] Scenario 1: The first information indicates the second and third functions.

[0145] In one possible scenario, the first device decomposes the first task into multiple functions. Then, based on the functions currently supported by the network device, the first device determines that the network device supports some of the multiple functions corresponding to the first task. At this time, the first device can indicate to the network device the functions supported by the network device among the multiple functions through the first information, and can also indicate to the network device the functions not supported by the network device among the multiple functions through the first information.

[0146] For example, illustrate whether a network device supports the functions associated with the first task. As an example, the functions supported by the network device are shown in Table 1. After decomposing the first task, five functions are obtained: Function A, Function B, Function C, Function D, and Function E. Option 1: If the specific function of Function A is the same as the function description of Function #1 in Table 1, the specific function of Function B is the same as the function description of Function #2 in Table 1, and the specific functions of Functions C, D, and E are all different from the function description of Function #3 in Table 1, then the terminal considers the network device to support Functions A and B among the multiple functions corresponding to the first task. Option 2: Based on the description in Option 1, for example, if the accuracy required to implement Function A is higher than any value in the accuracy range corresponding to Function #1, then the terminal considers the network device to not support Function A.

[0147] In this application, when a terminal instructs a network device to perform a third function, it can be understood that the terminal requests the network device to support the third function. For example, if the computational requirements for implementing the third function are large, the terminal may not be able to complete the computation. Therefore, the terminal can instruct the network device to perform the third function through the first information, so that the network device can execute the third function in the subsequent functional division of the first task, thereby avoiding the failure of the first task due to an inappropriate functional division method, which would affect the user experience.

[0148] In one possible implementation, if the network device supports the functions shown in Table 1, then the first information may include the function identifier of the second function and the function information of the third function. For example, the function information of the third function may be presented to the network device in a table format. For example, the function information and function identifier of the third function can be populated into Table 1, and the populated Table 1 can be sent to the network device to update the locally stored Table 1; alternatively, a new table can be created and sent to the network device to update the locally stored Table 1, where the new table includes the function information and function identifier of the third function.

[0149] Optionally, if the terminal also does not support the third function, it can indicate the specific function of the third function to the network device through the first information (for example, the first information includes the function description parameters of the third function).

[0150] Based on the above implementation method, the first information can also include the AI ​​model corresponding to the third function. The network device can deploy the corresponding AI model on the network device side, and after installation, the network device can also support the third function. Afterwards, the network device updates Table 1, that is, it adds the function information and function identifier of the third function to Table 1.

[0151] Based on the above implementation method, if the first information does not contain the AI ​​model corresponding to the third function, the network device can also obtain the AI ​​model corresponding to the third function from other devices based on the relevant information of the third function in the first information, deploy the corresponding AI model on the network device side, and the network device can also support the third function after installation. Afterwards, the network device updates Table 1 in its local storage, that is, it adds the function information and function identifier of the third function to Table 1.

[0152] Scenario 2: The first information indicates the second function.

[0153] In one possible scenario, the first device decomposes the first task into multiple functions. Then, based on the functions currently supported by the network device, the first device determines that the network device supports some or all of the functions corresponding to the first task. In this case, the first device can indicate to the network device the supported functions among the multiple functions corresponding to the first task via the first information, without needing to indicate the unsupported functions. In contrast, if the network device itself does not support the third function, and the first information does not indicate the third function, the network device will not acquire the AI ​​model corresponding to the third function, and the network device will still be unable to support the third function in subsequent processes. For how to indicate the second function, please refer to the description in scenario one; it will not be repeated here.

[0154] Scenario 3: The first information indicates the third function.

[0155] In one possible scenario, the first device decomposes the first task into multiple functions. Then, based on the functions currently supported by the network device, the first device determines that the network device does not support all of the multiple functions corresponding to the first task. The first device can then use first information to indicate to the network device the functions it expects the network device to support. How to indicate the third function is described in Case 1 and will not be repeated here.

[0156] Optionally, the first information is also used to indicate the execution order of the second and third functions. It can be understood that after aligning the model functions with the first device, the network device needs to perform functional division on the multiple functions corresponding to the first task, that is, to determine which functions of the multiple functions are executed by the terminal, network device, and server respectively. Therefore, the first device can also inform the network device of the execution order of the multiple functions.

[0157] Optionally, the method further includes:

[0158] S460, the terminal sends second information to the network device. Correspondingly, the network device receives the second information from the terminal. The second information indicates the functions supported by the terminal, which are included in both the second and third functions.

[0159] The functions supported by the terminal are included in the second and third functions. This can be understood as the functions indicated by the second information being included in the multiple functions obtained from the decomposition of the first task. In subsequent processes, the network device can perform function partitioning based on the second information. For example, the network device can know which functions the terminal supports and which it does not support based on the second information. Therefore, when partitioning functions, the network device will not assign functions that the terminal does not support. Another example is that the second information may include functional information about the functions supported by the terminal. If both the terminal and the network device support the image recognition function corresponding to the first task, but the terminal's image recognition function has a higher accuracy, the network device can consider implementing the image recognition function on the terminal side when partitioning functions. Based on the above description, it can be seen that the terminal sending the second information to the network device allows the network device and the terminal to align the functions that the terminal can perform, thus avoiding inappropriate function partitioning that could prevent the first task from being completed and affect the user experience.

[0160] Understandably, servers generally do not need to report the functions they support to network devices, as network devices assume that servers support all functions.

[0161] Optionally, the method further includes:

[0162] S470: The network device sends the first configuration information to the terminal. Correspondingly, the terminal receives the first configuration information from the network device.

[0163] The first configuration information indicates the available functions of the terminal, which are included in the second and third functions (similarly, the available functions are included in the multiple functions obtained from the first task decomposition).

[0164] It can be understood that the available functions of a terminal are those functions supported by the terminal.

[0165] It can also be understood that the available function is one of the functions associated with the first task that can be implemented on the terminal side, or that executing the available function on the terminal side can meet the performance requirements of the function.

[0166] The following example illustrates the first configuration information. For instance, after decomposing the first task, five functions are obtained: Function #1, Function #2, Function #3, Function #4, and Function #5. These five functions are then assigned to three devices (a terminal, a network device, and a server) for execution. Multiple allocation methods are possible. The network device can select several suitable allocation methods based on the currently acquired information and inform the terminal through the first configuration information. Afterward, the network device will only select one of the configured function allocation methods for function allocation. For example, the first configuration information includes two allocation methods: the first allocation method is that the terminal executes Function #1 and Function #2, the network device executes Function #3 and Function #4, and the server executes Function #5; the second allocation method is that the terminal executes Function #1, Function #2, and Function #3, the network device executes Function #4, and the server executes Function #5. For example, if the terminal supports functions #1 to #5, without sending the first configuration information, the terminal would need to reserve enough resources for the five functions to load the AI ​​models corresponding to functions #1 to #5 locally. However, in this method, since the network device informs the terminal in advance of the possible function division, the terminal only needs to reserve enough resources for functions #1, #2, and #3 to load the AI ​​models corresponding to functions #1 to #3, thereby saving terminal resources.

[0167] S430, the terminal obtains the calculation result of the first task.

[0168] It is understandable that the final calculation result of the first task needs to be sent to the terminal. The calculation result of the first task is determined based on the AI ​​models of the second and third functions; that is, the calculation result of the first task is determined based on the AI ​​models of the multiple functions obtained from the decomposition of the first task.

[0169] In one possible scenario, if the network device supports all functions associated with the first task, then the third function does not exist. In this case, the calculation result of the first task is determined based on the AI ​​model of the second function. In another possible scenario, if the network device does not support all functions associated with the first task, then the second function does not exist. In this case, the calculation result of the first task is determined based on the AI ​​model of the third function.

[0170] Optionally, the method also includes S440 and S450:

[0171] S440, the server obtains the first calculation result, which is determined based on the AI ​​model corresponding to the second function and / or the third function.

[0172] It can be understood that after the network device divides the first task into functional parts, if the server needs to perform a certain function, then the server will perform that function. The first calculation result is determined based on the AI ​​model corresponding to the second and / or third function. This can be understood as the first calculation result being determined based on the AI ​​model corresponding to the function that the server needs to perform among the multiple functions obtained from the decomposition of the first task.

[0173] S450, the network device obtains the second calculation result, which is determined based on the AI ​​model corresponding to the second function and / or the third function.

[0174] It can be understood that after the network device divides the first task into functions, if the network device needs to perform a certain function, then the network device will perform that function. The second calculation result is determined based on the AI ​​model corresponding to the second and / or third functions. This can be understood as the second calculation result being determined based on the AI ​​model corresponding to the function that the network device needs to perform among the multiple functions obtained from the decomposition of the first task.

[0175] It can also be understood that if a network device needs to perform a third function, the AI ​​model corresponding to the third function has already been obtained before the network device performs the third function. The method of obtaining this model is described in S420 and will not be repeated here.

[0176] In the above technical solution, the terminal can align with the network device to complete the first task based on the functions supported by the acquired network device, enabling distributed reasoning of the first task (i.e., different devices implementing different functions of the first task) between the terminal and the network device. Based on this method, functions limited by the terminal's capabilities can be implemented on the network device side, thereby improving the user experience.

[0177] It can be understood that Figure 4 can be viewed as the network device establishing a service with the terminal (or server) before performing functional division of the first task (the process corresponding to Figure 4 can be called the service establishment phase). After that, the network device can perform functional division of the first task based on the information interacted with the terminal. Each device executes the function associated with the first task based on the functional division result, thereby obtaining the calculation result of the first task (this process can be called the service execution phase). The steps of the service execution phase will be described in detail below with reference to Figure 5 as an example.

[0178] Figure 5 is a schematic flowchart of a communication method 500 provided in this application. The method includes the following steps.

[0179] S510, the terminal sends a second request message to the network device, which requests the network device to perform functional division of the first task. Correspondingly, the network device receives the second request message.

[0180] Optionally, the second request message includes information related to the terminal's computing resources. For example, the information related to the terminal's computing resources indicates the amount of video memory, RAM, and computing power that the terminal can support.

[0181] S520 network devices divide the functions of the primary task.

[0182] In this application, the functional division of the first task can also be described as the task arrangement of the first task. Correspondingly, the functional division result of the first task can also be described as the task arrangement result of the first task.

[0183] It's understandable that when network devices divide the functions of the first task, they need to consider one or more factors, such as the current air interface status, the computing resources of network devices and terminals, and the quality of service (QoS) of the first task, based on the specific functions involved. For example, the dimensions of the output data may differ after executing different functions, and the air interface status will affect the end-to-end QoS. Furthermore, different functions require different computing resources; considering the computing resources of terminals and network devices is crucial for better function division. Also, QoS requirements include end-to-end service latency and accuracy; different devices executing different functions will have different end-to-end service latency and success rates, therefore the QoS of the first task will also affect function division.

[0184] Let's illustrate with examples. For instance, the first task is to instruct an intelligent robot to pick up a package. This task can be broken down into two functions: understanding the visual environment and understanding human commands (i.e., language understanding). For example, the robot itself can understand the visual environment but not the language understanding function. Therefore, when dividing the task, the network device cannot assign the language understanding function to the terminal side. Another example is that the QoS requirements of the first task include end-to-end service latency. It's understandable that the allocation methods for guaranteeing 200ms of service latency differ from those for guaranteeing 1s of service latency. If the visual environment understanding function is assigned to the terminal side, the corresponding service latency would be 300ms, while the QoS requirement is 200ms. Therefore, placing the visual environment understanding function on the terminal side would result in unmet service quality requirements. Furthermore, the dimensions of the output data for different functions differ, the number of quantization bits may differ, and the data volume will also differ. When communication conditions are poor (i.e., poor air interface conditions), a function allocation method with lower communication transmission volume should be chosen; conversely, a allocation method with higher communication transmission volume but lower computational latency can be chosen.

[0185] In S530, the network device sends the function partitioning results to the terminal. Correspondingly, the terminal receives the function partitioning results from the network device.

[0186] For example, network devices can send function partitioning results to terminals via downlink control information (DCI) or medium access control (MAC) messages.

[0187] It is understandable that if the partitioning result indicates that the server still needs to calculate the function of the first task association, then the network device also needs to send the partitioning result to the server.

[0188] In S540, terminals, network devices, and servers execute corresponding functions based on the functional partitioning results.

[0189] For example, the first task can be decomposed into function #1, function #2, and function #3. The function division result indicates that function #1 is implemented on the terminal side, function #2 is implemented on the network device side, and function #3 is implemented on the server side. The execution order of the three functions is function #1, function #2, and function #3, respectively. The network device sends the function division result to the terminal and the server respectively. Then, the terminal performs calculations on the terminal side based on the AI ​​model #1 corresponding to function #1 and sends the output result #1 of AI model #1 to the network device. The network device performs calculations on the network device side based on the output result #1 and the AI ​​model #2 corresponding to function #2 and sends the output result #2 of AI model #2 to the server. Then, the server performs calculations on the server side based on the output result #2 and the AI ​​model #3 corresponding to function #3 and returns the output result #3 of AI model #3 (i.e., an example of the first calculation result) to the terminal (correspondingly, in S430, the terminal obtains the calculation result of the first task).

[0190] It is understood that the steps in the above figures are merely illustrative and are not intended to be strictly limited. Furthermore, the sequence numbers of the processes described above do not imply a specific order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0191] It is also understood that some optional features in the various embodiments of this application may not depend on other features in some scenarios, or may be combined with other features in some scenarios, without limitation.

[0192] It is also understood that, in the above-described method embodiments, the methods and operations implemented by the device (terminal or network device) can also be implemented by components of the device (e.g., chips or circuits), without limitation.

[0193] The method embodiments provided in this application have been described in detail above with reference to Figures 1 to 5. The apparatus embodiments of this application will be described below with reference to Figures 6 and 7. It is understood that, in order to implement the functions in the above embodiments, the apparatuses in Figures 6 and 7 include hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should readily recognize that, based on the units and method steps of the various examples described in conjunction with the embodiments disclosed in this application, this application can be implemented in hardware or a combination of hardware and computer software. It is understood that the technical features described in the above method embodiments are also applicable to the following apparatus embodiments.

[0194] Figures 6 and 7 are schematic diagrams of possible apparatus structures provided in embodiments of this application. These apparatuses can be used to implement the functions of the terminal or network device in the above method embodiments, and thus can also achieve the beneficial effects of the above method embodiments.

[0195] Figure 6 is a schematic block diagram of a communication device 600 provided in an embodiment of this application. As shown in Figure 6, the communication device 600 may include modules or units for implementing the methods described above. In one possible design, the communication device 600 includes a processing unit 602 and a communication unit 603. Optionally, the communication device 600 may further include a storage unit 601 for storing device program code and / or data.

[0196] The communication device 600 can be a terminal-side device as described in the above embodiments, such as a terminal or a communication module in a terminal, or a circuit or chip in a terminal that is responsible for communication functions.

[0197] For example, in one embodiment, the processing unit 602 is configured to: acquire a first function supported by the network device; the communication unit 603 is configured to: send first information to the network device, the first information indicating a second function and / or a third function, the second function and the third function being associated with a first task, the first function including the second function but not including the third function; the processing unit 602 is further configured to: acquire the calculation result of the first task, the calculation result being determined based on the AI ​​model of the second function and the third function.

[0198] In one possible design, the first function is defined by function information, and the first function corresponds to a function identifier. The function information indicates one or more of the following of the first function: the dimension of the input information, the modality of the input information, the description of the input information, the dimension of the output information, the modality of the output information, the description of the output information, accuracy, computation latency, memory usage, computing power requirements, and function description.

[0199] In one possible design, the first information is also used to indicate the execution order of the second function and the third function.

[0200] In one possible design, the communication unit 603 is further configured to: send second information to the network device, the second information indicating the functions supported by the terminal, the functions supported by the terminal being included in the second function and the third function.

[0201] In one possible design, the communication unit 603 is further configured to: receive first configuration information from the network device, the first configuration information indicating the available functions of the terminal, the available functions being included in the second function and the third function, and the processing unit 602 is specifically configured to: obtain the calculation result of the first task based on the AI ​​model corresponding to the available functions.

[0202] In one possible design, the communication unit 603 is further configured to: receive third information from the network device; the processing unit 602 is specifically configured to: obtain the first function based on the third information.

[0203] In one possible design, the communication unit 603 is further configured to: send a first request message to the network device, the first request message being used to request the first function.

[0204] In one possible design, when the communication device 600 is a terminal or a communication module within a terminal, the function of the processing unit 602 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 603 can be implemented by transceiver circuitry.

[0205] In one possible design, when the communication device 600 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 602 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 603 can be implemented by an interface circuit or data transceiver circuit on the aforementioned chip.

[0206] In one possible design, when the communication device 600 is a terminal or a processing module within a terminal, the functionality of the processing unit 602 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. The functionality of the communication unit 603 can be implemented by transceiver circuitry.

[0207] In one possible design, when the communication device 600 is a circuit or chip in the terminal responsible for processing functions, such as a GPU or a system-on-a-chip (SoC) or SIP chip containing a GPU, the function of the processing unit 602 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 603 can be implemented by interface circuitry or data transceiver circuitry on the aforementioned chip.

[0208] 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.

[0209] 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.

[0210] In one example, storage unit 601 may include random access memory, flash memory, read-only memory, programmable read-only memory or electrically erasable programmable memory and / or registers, etc.

[0211] Figure 7 is a schematic diagram of the structure of a terminal 1000 provided in an embodiment of this application. The terminal 1000 corresponds to the terminal shown in Figure 1 and is used to implement the operation of the terminal in the above embodiments. As shown in Figure 7, the terminal includes: one or more antennas 1010, a radio frequency processing system 1020, and a processor system 1030.

[0212] In the downlink or sidelink direction, the RF processing system 1020 receives RF signals through the antenna 1010 and sends the RF-processed signals to the processor system 1030 for further processing. In the uplink or sidelink direction, the processor system 1030 processes the terminal-side information and sends it to the RF processing system 1020, which then processes the signal and transmits it through the antenna 1010.

[0213] In one example, the RF processing system 1020 serves as the communication interface for external communication of the terminal and may include an RF front end (RFFE) 1021 and an RF transceiver 1022. The RFFE 1021 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 1021 can be a circuit system composed of multiple discrete devices or integrated into one or more chips. The RF transceiver 1022 processes the RF signals received by the RFFE into baseband / IF signals for further processing by the processor system 1030, and processes the baseband / IF signals provided by the processor system 1030 into RF signals for transmission to the RFFE 1021. The baseband / IF signals transmitted between the RF transceiver 1022 and the processor system 1030 can be digital or analog signals. The radio frequency transceiver 1022 can be implemented by one or more chips, which are commonly referred to as radio frequency chips (RFICs).

[0214] In one example, processor system 1030 may include one or more processors for processing signals and executing one or more communication protocols. Optionally, processor system 1030 may also include memory 1036. In one example, the one or more processors include at least one baseband processor 1031 (also known as a modem processor). Memory 1036 is used to store data and / or computer program instructions. Optionally, processor system 1030 may also include one or more application processors 1032 for implementing processing of the terminal operating system and application layer. Application processor 1032 may include, for example, a GPU. Optionally, processor system 1030 may also include one or more of a voice subsystem 1033, a multimedia subsystem 1034, or an interface circuit 1035. The voice subsystem 1033 is used to process voice signals, the multimedia subsystem 1034 is used to handle multimedia-related operations, such as video encoding / decoding, image processing, etc., and the interface circuit 1035 is used to implement communication with other terminal components, such as a display 1040, an input device 1050, memory 1060, etc. The aforementioned components in the processor system 1030 can communicate with each other via a bus or communication interface circuit.

[0215] In one example, the processor system 1030 can be packaged as a single processor chip, such as a SoC chip or a SIP chip. In another example, the processor system 1030 can be a system composed of multiple chips, for example, the baseband processor 1031 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.

[0216] In one example, memory 1036 can be on-chip memory, i.e., located on the processor system 1030 chip. In another example, memory 1060 can be off-chip memory, i.e. located outside the processor system 1030 chip.

[0217] In one example, the baseband processor 1031 may include one or more processor cores 10311 and interface circuitry 10314. The one or more processor cores 10311 are used to process signals and execute one or more communication protocols. Optionally, the baseband processor 1031 may also include a memory 10312 for storing at least a portion of the corresponding computer program instructions and / or data. In one example, the one or more processor cores 10311 execute the computer program instructions stored in the memory 10312 to implement the relevant operations in the above method embodiments (e.g., functional decomposition of a first task, or, for example, obtaining functions supported by the network device based on third information from the network device). In this disclosure, memory 10312 is used to store corresponding computer program instructions and / or data. This can mean that memory 10312 stores all corresponding computer program instructions and / or data for execution by processor core 10311; or it can mean that memory 10312 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 10311. Memory 10312 can store different portions of computer program instructions and / or data multiple times for execution by processor core 10311 to implement the relevant operations in the above method embodiments. Interface circuit 10314 serves as a communication interface for communication with other components, such as transmitting signals with radio frequency processing system 1020, communicating with other subsystems and related components of processor system 1030 via bus, such as transmitting data control signals with application processor 1032, and transmitting data or computer program instructions with memory 1036 or memory 1060. Optionally, in order to reduce the load on the processor core, a baseband signal processing circuit 10313 can be set to perform at least some baseband signal processing, including one or more of signal demodulation, modulation, encoding or decoding.

[0218] In one example, the communication device provided in this application may be a terminal 1000, a communication module including a processor system 1030 and a radio frequency system 1020, or a baseband processor 1031.

[0219] 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), artificial intelligence processor (AI processor), or neural processing unit (NPU).

[0220] 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 in non-volatile memory, such as at least a portion of the aforementioned memory 1060 (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 1036 and / or memory 10312 (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.

[0221] It should also be noted that the memory described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0222] In one example, the RF transceiver 1022 and the RF front-end 1021 can also be packaged in a single chip. In another example, the RF transceiver 1022, the RF front-end 1021, and the baseband processor 1031 can also be packaged in a single chip. Furthermore, this application also provides a computer-readable storage medium storing computer instructions that, when executed on a computer, cause operations and / or processes performed by a terminal or network device in the various method embodiments of this application to be executed.

[0223] This application also provides a computer program product, which includes computer program code or instructions. When the computer program code or instructions are run on a computer, the operations and / or processes performed by a terminal or network device in the various method embodiments of this application are executed.

[0224] Furthermore, this application also provides a chip including a processor. A memory for storing a computer program is provided independently of the chip, and the processor is used to execute the computer program stored in the memory, such that operations and / or processes performed by a terminal or network device in any method embodiment are performed.

[0225] Furthermore, the chip may also include a communication interface. The communication interface may be an input / output interface or an interface circuit, etc. Furthermore, the chip may also include a memory.

[0226] In addition, this application also provides a communication system, including at least one of the terminal, network device and server in the embodiments of this application.

[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, create means for implementing the functions specified in one or more blocks of the flowchart illustrations and / or one or more blocks of the block diagrams.

[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 that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0230] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable apparatus, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[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 communication method characterized by comprising: The method comprises: obtaining a first function supported by a network device; sending first information to the network device, the first information indicating a second function and / or a third function, the second function and the third function being associated with a first task, the first function including the second function and not including the third function; obtaining a calculation result of the first task, the calculation result being determined based on an artificial intelligence (AI) model of the second function and the third function.

2. The method of claim 1, wherein: the first function is defined by function information, and the first function corresponds to a function identifier, the function information indicating one or more of the following of the first function: dimension of input information, modality of input information, description of input information, dimension of output information, modality of output information, description of output information, accuracy, calculation delay, video memory occupation, computing power requirement, and function description. the first information is further used to indicate an execution order of the second function and the third function.

3. The method according to claim 1 or 2, characterized in that, The method further comprises:

4. The method according to any one of claims 1 to 3, characterized in that, sending second information to the network device, the second information indicating a function supported by a terminal, the function supported by the terminal being included in the second function and the third function. The method further comprises:

5. The method according to any one of claims 1 to 4, characterized in that, receiving first configuration information from the network device, the first configuration information indicating available functions of a terminal, the available functions being included in the second function and the third function, the obtaining of the calculation result of the first task comprises: obtaining the calculation result of the first task based on an AI model corresponding to the available functions. The obtaining of the first function supported by the network device comprises:

6. The method according to any one of claims 1 to 5, characterized in that, receiving third information from the network device; obtaining the first function based on the third information. The method further comprises:

7. The method of claim 6, wherein, sending a first request message to the network device, the first request message being used to request the first function. The apparatus comprises:

8. A communication device, characterized by a processing unit, configured to obtain a first function supported by a network device; a communication unit, configured to send first information to the network device, the first information indicating a second function and / or a third function, the second function and the third function being associated with a first task, the first function including the second function and not including the third function; the processing unit is further configured to obtain a calculation result of the first task, the calculation result being determined based on an artificial intelligence (AI) model of the second function and the third function.

9. The apparatus of claim 8, wherein: the first function is defined by function information, and the first function corresponds to a function identifier, the function information indicating one or more of the following of the first function: dimension of input information, modality of input information, description of input information, dimension of output information, modality of output information, description of output information, accuracy, calculation delay, video memory occupation, computing power requirement, and function description. the first information is further used to indicate an execution order of the second function and the third function.

11. The apparatus of any one of claims 8 to 10, wherein:

10. The apparatus of claim 8 or 9, wherein, ​ ​ The communication unit is further configured to send second information to the network device, the second information indicating functions supported by the terminal, the functions supported by the terminal being included in the second function and the third function.

12. The apparatus of any one of claims 8-11, wherein, The communication unit is further configured to receive first configuration information from the network device, the first configuration information indicating available functions of the terminal, the available functions being included in the second function and the third function, The processing unit is configured to obtain a calculation result of the first task, including: The processing unit is configured to obtain the calculation result of the first task based on the AI model corresponding to the available function.

13. The apparatus of any one of claims 8-12, wherein, The communication unit is further configured to receive third information from the network device. The processing unit is configured to obtain a first function supported by the network device, including: The processing unit is configured to obtain the first function based on the third information.

14. The apparatus of claim 13, wherein, The communication unit is further configured to send a first request message to the network device, the first request message being used to request the first function.

15. A computer-readable storage medium, characterized in that, The storage medium has stored therein a computer program or instructions, which, when executed, cause the method of any one of claims 1-7 to be implemented.

16. A computer program product, characterised in that, The computer program, when executed, causes the method of any one of claims 1-7 to be implemented.

17. A communications device, characterized by The apparatus includes units for implementing the method of any one of claims 1-7.

18. A communications device, characterized by The apparatus includes an interface circuit and one or more processors coupled with a memory, the memory being configured to store computer programs or instructions, which, when executed by the one or more processors, cause the apparatus to implement the method of any one of claims 1-7.

19. The apparatus of claim 18, wherein, The interface circuit is configured to implement communication functions within the apparatus and / or between the apparatus and other apparatuses or components. The interface circuit is configured to implement communication functions within the apparatus and / or between the apparatus and other apparatuses or components.

Images