Artificial intelligence (AI) unit configuration method and device and communication equipment

Abstract

The invention discloses an artificial intelligence AI unit configuration method and device and communication equipment, and belongs to the technical field of communication, and the AI unit configuration method comprises the steps that a first network function sends first configuration information to a first node; wherein the first node comprises a node related to a first AI unit; the first configuration information is used for configuring at least one of node update information, data interaction update information, input data time sequence change information and output data time sequence change information related to the first AI unit; the first AI unit comprises an AI unit generated based on a non-AI function, or the first AI unit comprises M AI units generated based on N second AI units deployed at a second node, N and M are positive integers, and N is not equal to M.

Description

Technical Field

[0001] The present application belongs to the field of communication technology, and specifically relates to an artificial intelligence (AI) unit configuration method, device and communication equipment. Background Art

[0002] In mobile communication systems, more and more use cases are beginning to incorporate AI. The 3rd Generation Partnership Project (3GPP) Radio Access Network (RAN) physical layer working group discussed, for example: AI-based channel state information (CSI) prediction and feedback compression, AI-based beam management, and AI-based positioning at the physical layer. The RAN3 working group discussed AI-based energy saving, AI-based load balancing, etc. In the future, there will be more use cases incorporating AI in mobile communication systems.

[0003] However, as wireless environments and communication system requirements continue to change over time, AI use cases and AI units also need to evolve continuously to adapt to communication systems. In related technologies, research is underway on how to automate the evolution of AI units and AI use cases using a self-evolutionary approach.

[0004] In the related art, during the AI ​​unit update process, in order to facilitate the use of the updated AI unit, at least one of the following conditions needs to be met:

[0005] 1) The deployment node of the updated AI unit remains unchanged. For example, multiple AI units are stored in a node, where the model structure and parameters of each AI unit are fixed. Then, by instructing the node to switch the AI ​​unit, the deployment nodes of the AI ​​units before and after the switch do not change;

[0006] 2) The model structure of the updated AI unit remains unchanged. For example, if the parameters of the AI ​​unit are updated by instructing the deployment node of the AI ​​unit, the model structure of the AI ​​unit does not change.

[0007] However, based on the above limitations, the flexibility of the AI ​​unit update process will be reduced or the amount of interaction data between nodes will increase. For example, assuming that an AI unit is deployed on node A, it needs to obtain input data from node B. If the updated AI unit is still deployed on node A, then the node needs to interact with node B to obtain input data. This limits the data interaction between nodes A and B when the AI ​​unit obtains input data. Summary of the invention

[0008] The embodiments of the present application provide an AI unit configuration method, apparatus, and communication device, which can change the deployment nodes or model structure of the AI ​​unit through the AI ​​unit self-evolution process, and after the deployment nodes or model structure of the AI ​​unit caused by the AI ​​unit self-evolution changes, the nodes related to the changed AI unit can be instructed through the first configuration information to collaboratively implement the data collection, model reasoning, application of model reasoning results, and other processes of the first AI unit to realize the AI ​​function of the updated AI unit.

[0009] In a first aspect, a method for configuring an AI unit is provided, the method comprising:

[0010] The first network function sends first configuration information to the first node;

[0011] The first node includes a node related to a first AI unit; the first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to the first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on a second node, where N and M are positive integers, and N is not equal to M.

[0012] In a second aspect, another AI unit configuration method is provided, the method comprising:

[0013] The first node receives first configuration information from the first network function, wherein the first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to the first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on the second node, where N and M are positive integers, and N is not equal to M;

[0014] The first node performs a first operation according to the first configuration information, where the first operation includes at least one of the following:

[0015] Sending updated input data to at least one third node where the first AI unit is deployed;

[0016] Obtain input data of the first AI unit from the updated input node;

[0017] Based on the first AI unit, perform model reasoning on the updated input data to obtain first output data or model reasoning results;

[0018] Sending second output data to at least one node where a first AI unit component is deployed, wherein the second output data includes output data of a second AI unit component, and the first AI unit includes at least one of the first AI unit component and the second AI unit component;

[0019] Sending the model inference result of the first AI unit to the fourth node;

[0020] Obtain the model inference result of the first AI unit from the third node.

[0021] In a third aspect, an AI unit configuration device is provided for a first network function, the device comprising:

[0022] A sending module, used to send first configuration information to the first node;

[0023] The first node includes a node related to a first AI unit; the first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to the first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on a second node, where N and M are positive integers, and N is not equal to M.

[0024] In a fourth aspect, an AI unit configuration device is provided, for use in a first node, the device comprising:

[0025] a receiving module, configured to receive first configuration information from a first network function, wherein the first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to a first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on a second node, where N and M are positive integers, and N is not equal to M;

[0026] An execution module is configured to execute a first operation according to the first configuration information, where the first operation includes at least one of the following:

[0027] Sending updated input data to at least one third node where the first AI unit is deployed;

[0028] Obtain input data of the first AI unit from the updated input node;

[0029] Based on the first AI unit, perform model reasoning on the updated input data to obtain first output data or model reasoning results;

[0030] Sending second output data to at least one node where a first AI unit component is deployed, wherein the second output data includes output data of a second AI unit component, and the first AI unit includes at least one of the first AI unit component and the second AI unit component;

[0031] Sending the model inference result of the first AI unit to the fourth node;

[0032] Obtain the model inference result of the first AI unit from the third node.

[0033] In a fifth aspect, a terminal is provided, comprising a processor and a memory, wherein the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the steps of the method described in the second aspect are implemented.

[0034] In a sixth aspect, a terminal is provided, including a processor and a communication interface, wherein the communication interface is used to receive first configuration information from a first network function, wherein the first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to a first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on a second node, N and M are positive integers, and N is not equal to M; the processor is used to perform a first operation according to the first configuration information, and the first operation includes at least one of the following:

[0035] Sending updated input data to at least one third node where the first AI unit is deployed;

[0036] Obtain input data of the first AI unit from the updated input node;

[0037] Based on the first AI unit, perform model reasoning on the updated input data to obtain first output data or model reasoning results;

[0038] Sending second output data to at least one node where a first AI unit component is deployed, wherein the second output data includes output data of a second AI unit component, and the first AI unit includes at least one of the first AI unit component and the second AI unit component;

[0039] Sending the model inference result of the first AI unit to the fourth node;

[0040] Obtain the model inference result of the first AI unit from the third node.

[0041] In the seventh aspect, a network side device is provided, which includes a processor and a memory, wherein the memory stores programs or instructions that can be run on the processor, and when the program or instructions are executed by the processor, the steps of the method described in the first aspect or the second aspect are implemented.

[0042] In an eighth aspect, a network side device is provided, including a processor and a communication interface;

[0043] In some embodiments, the communication interface is used to send first configuration information to the first node;

[0044] The first node includes a node related to the first AI unit; the first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to the first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on the second node, where N and M are positive integers, and N is not equal to M;

[0045] In some other embodiments, the communication interface is used to receive first configuration information from a first network function, wherein the first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to a first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on a second node, where N and M are positive integers and N is not equal to M; the processor is used to perform a first operation according to the first configuration information, wherein the first operation includes at least one of the following:

[0046] Sending updated input data to at least one third node where the first AI unit is deployed;

[0047] Obtain input data of the first AI unit from the updated input node;

[0048] Based on the first AI unit, perform model reasoning on the updated input data to obtain first output data or model reasoning results;

[0049] Sending second output data to at least one node where a first AI unit component is deployed, wherein the second output data includes output data of a second AI unit component, and the first AI unit includes at least one of the first AI unit component and the second AI unit component;

[0050] Sending the model inference result of the first AI unit to the fourth node;

[0051] Obtain the model inference result of the first AI unit from the third node.

[0052] In a ninth aspect, a readable storage medium is provided, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the method described in the first aspect are implemented, or the steps of the method described in the second aspect are implemented.

[0053] In the tenth aspect, a wireless communication system is provided, comprising: a terminal and a network side device, wherein the terminal can be used to execute the steps of the method described in the second aspect, and the network side device can be used to execute the steps of the method described in the first aspect, or the first aspect and the second aspect.

[0054] In the eleventh aspect, a chip is provided, comprising a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run a program or instruction to implement the method described in the first aspect, or to implement the method described in the second aspect.

[0055] In the twelfth aspect, a computer program / program product is provided, wherein the computer program / program product is stored in a storage medium, and the program / program product is executed by at least one processor to implement the steps of the AI ​​unit configuration method as described in the first aspect or the second aspect.

[0056] In the embodiment of the present application, during the evolution of the AI ​​unit, the deployment nodes and model structure of the evolved first AI unit are not restricted, and the first configuration information is sent to the nodes related to the first AI unit through the first network function, so that each node related to the first AI unit can obtain at least one of the node update information, data interaction update information, input data timing change information and output data timing change information related to the first AI unit according to the first configuration information received by each node. In the scenario where the deployment of the AI ​​unit or the data interaction between nodes changes due to the generation of the first AI unit based on a non-AI function, or the second AI units deployed on at least two second nodes are merged into the first AI unit through function fusion or model fusion technology, or the second AI unit deployed on one second node is split into at least two first AI units through AI unit self-evolution, the first configuration information of the embodiment of the present application enables the nodes related to the first AI unit to collaboratively implement the data collection, model reasoning, application of model reasoning results, model monitoring and other processes of the first AI unit according to the first configuration information, so that the AI ​​unit evolution is more flexible and does not increase the amount of data interaction between nodes. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 is a structural diagram of a wireless communication system to which the embodiments of the present application can be applied;

[0058] Figure 2 is a flow chart of an AI unit configuration method provided in an embodiment of the present application;

[0059] Figure 3a is a schematic diagram of a second AI unit deployed on at least two second nodes in an embodiment of the present application;

[0060] Figure 3b is a schematic diagram of deployment scheme 1 of the first AI unit in an embodiment of the present application;

[0061] Figure 3c is a schematic diagram of deployment scheme 2 of the first AI unit in an embodiment of the present application;

[0062] Figure 3d is a schematic diagram of deployment scheme 3 of the first AI unit in an embodiment of the present application;

[0063] Figure 4 is a flowchart of another AI unit configuration method provided in an embodiment of the present application;

[0064] Figure 5 It is a structural diagram of an AI unit configuration device provided in an embodiment of the present application;

[0065] Figure 6 is a structural diagram of another AI unit configuration device provided in an embodiment of the present application;

[0066] Figure 7 It is a structural diagram of a communication device provided in an embodiment of the present application;

[0067] Figure 8 is a schematic diagram of the structure of a terminal provided in an embodiment of the present application;

[0068] Fig. 9 It is a structural diagram of a network side device provided in an embodiment of the present application;

[0069] Fig.10 It is a structural diagram of another network side device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0070] The following will be combined with the drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of this application.

[0071] The terms "first", "second", etc. of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the terms used in this way are interchangeable where appropriate, so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of one type, and the number of objects is not limited, for example, the first object can be one or more. In addition, "or" in the present application represents at least one of the connected objects. For example, "A or B" covers three schemes, namely, Scheme 1: including A but not including B; Scheme 2: including B but not including A; Scheme 3: including both A and B. The character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0072] The term "indication" in this application can be a direct indication (or explicit indication) or an indirect indication (or implicit indication). A direct indication can be understood as the sender explicitly informing the receiver of specific information, operations to be performed, or request results in the sent indication; an indirect indication can be understood as the receiver determining the corresponding information according to the indication sent by the sender, or making a judgment and determining the operation to be performed or the request result according to the judgment result.

[0073] It is worth noting that the technology described in the embodiments of the present application is not limited to the Long Term Evolution (LTE) / LTE-Advanced (LTE-A) system, but can also be used in other wireless communication systems, such as Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), Orthogonal Frequency Division Multiple Access (OFDMA), Single-carrier Frequency-Division Multiple Access (SC-FDMA) or other systems. The terms "system" and "network" in the embodiments of the present application are often used interchangeably, and the described technology can be used for the systems and radio technologies mentioned above as well as for other systems and radio technologies. The following description describes a New Radio (NR) system for illustrative purposes, and NR terminology is used in most of the following descriptions, but these technologies can also be applied to systems other than NR systems, such as the 6th generation (6 thGeneration, 6G) communication system.

[0074] Figure 1A block diagram of a wireless communication system applicable to an embodiment of the present application is shown. The wireless communication system includes a terminal 11 and a network side device 12. Among them, the terminal 11 can be a mobile phone, a tablet computer (Tablet Personal Computer), a laptop computer (Laptop Computer), a notebook computer, a personal digital assistant (Personal Digital Assistant, PDA), a handheld computer, a netbook, an ultra-mobile personal computer (Ultra-mobile Personal Computer, UMPC), a mobile Internet device (Mobile Internet Device, MID), an augmented reality (Augmented Reality, AR), a virtual reality (Virtual Reality, VR) device, a robot, a wearable device (Wearable Device), an aircraft (flight vehicle), a vehicle-mounted device (Vehicle User Equipment, VUE), a ship-mounted device, a pedestrian terminal (Pedestrian User Equipment, PUE), a smart home (home appliances with wireless communication functions, such as refrigerators, televisions, washing machines or furniture, etc.), a game console, a personal computer (Personal Computer, PC), a teller machine or a self-service machine and other terminal side devices. Wearable devices include: smart watches, smart bracelets, smart headphones, smart glasses, smart jewelry (smart bracelets, smart bracelets, smart rings, smart necklaces, smart anklets, smart anklets, etc.), smart wristbands, smart clothing, etc. Among them, the vehicle-mounted device can also be called a vehicle-mounted terminal, a vehicle-mounted controller, a vehicle-mounted module, a vehicle-mounted component, a vehicle-mounted chip or a vehicle-mounted unit, etc. It should be noted that the specific type of the terminal 11 is not limited in the embodiment of the present application. The network side device 12 may include an access network device or a core network device, wherein the access network device may also be referred to as a radio access network (Radio Access Network, RAN) device, a radio access network function or a radio access network unit. The access network device may include a base station, a wireless local area network (Wireless Local Area Network, WLAN) access point (Access Point, AS) or a wireless fidelity (Wireless Fidelity, WiFi) node, etc.Among them, the base station may be referred to as a Node B (NB), an evolved Node B (eNB), a next generation Node B (gNB), a New Radio Node B (NR Node B), an access point, a Relay Base Station (RBS), a Serving Base Station (SBS), a Base Transceiver Station (BTS), a radio base station, a radio transceiver, a Basic Service Set (BSS), an Extended Service Set (ESS), a Home Node B (HNB), a Home Evolved Node B (home evolved Node B), a Transmission Reception Point (TRP) or other appropriate terms in the relevant field. As long as the same technical effect is achieved, the base station is not limited to specific technical vocabulary. It should be noted that in the embodiment of the present application, only the base station in the NR system is used as an example for introduction, and the specific type of the base station is not limited.

[0075] The core network device may include, but is not limited to, at least one of the following: a core network node, a core network function, a mobility management entity (Mobility Management Entity, MME), an access mobility management function (Access and Mobility Management Function, AMF), a session management function (Session Management Function, SMF), a user plane function (User Plane Function, UPF), a policy control function (Policy Control Function, PCF), a policy and charging rules function unit (Policy and Charging Rules Function, PCRF), an edge application service discovery function (Edge Application Server Discovery Function, EASDF), a unified data management (Unified Data Management, UDM), a unified data storage (Unified Data Repository, UDR), a home user server (Home Subscriber Server, HSS), a centralized network configuration (CNC), a network storage function (Network Repository Function, NRF), a network exposure function (Network Exposure Function, NEF), a local NEF (Local NEF, or L-NEF), and a binding support function (Binding Support Function, BSF), application function (Application Function, AF), etc. It should be noted that in the embodiments of the present application, only the core network device in the NR system is introduced as an example, and the specific type of the core network device is not limited. It should be noted that in the embodiments of the present application, only the core network device in the NR system is introduced as an example, and the specific type of the core network device is not limited.

[0076] In order to facilitate the description of the AI ​​unit configuration method provided in the embodiment of the present application, the following nouns or terms in the embodiment of the present application are first explained:

[0077] 1) AI unit, which may also be referred to as AI model, machine learning (ML) model, ML unit, AI structure, AI function, AI characteristic, neural network, neural network function, neural network function, etc., or the AI ​​unit may also refer to a processing unit that can implement AI-related algorithms, formulas, processing procedures, capabilities, etc., or the AI ​​unit may be a processing method, algorithm, function, module or unit for a specific data set, or the AI ​​unit may be a processing method, algorithm, function, module or unit running on AI / ML related hardware such as a graphics processor (GPU), a neural network processor (NPU), a tensor processor (TPU), an application specific integrated circuit (ASIC), etc., and the embodiments of the present application do not specifically limit this.

[0078] Optionally, the specific data set includes at least one of input data and output data of the AI ​​unit.

[0079] 2) The identifier of the AI ​​unit may be an AI model identifier, an AI structure identifier, an AI algorithm identifier, or an identifier of a specific data set associated with the AI ​​unit, or an identifier of a specific scenario, environment, channel feature, or device related to the AI / ML, or an identifier of a function, feature, capability, or module related to the AI / ML. The embodiments of the present application do not specifically limit this.

[0080] 3) An input node of an AI unit may represent a node used to provide input data to the AI ​​unit.

[0081] 4) An output node of an AI unit may represent a node for receiving output data of the AI ​​unit.

[0082] 5) The first network function may be at least one of an AI unit management function, an AI unit control function, and an algorithm control function, which is not specifically limited here.

[0083] In some implementations, the entity device where the first network function is located may include at least one of a base station and a core network device, wherein one core network device may correspond to one core network function. Alternatively, the first network function may also be a network function independent of the cellular network system.

[0084] 6) The first node may include any node related to the first AI unit, and the number of the first nodes may be one or more. For example, the first node includes at least one of the following: a node on which the first AI unit is deployed, a node for providing input data to the first AI unit, a node for obtaining output data or model inference results of the first AI unit, a deployment node of an original AI unit for fusing to obtain the first AI unit, etc., which is not specifically limited here.

[0085] In some implementations, the entity device where the first node is located may include at least one of a UE, a base station, and a core network device.

[0086] It is worth noting that when the first node and the first network function are co-located in the same physical device, the interaction between the first node and the first network function is internally implemented; or, when at least two first nodes are co-located in the same physical device, the interaction between the at least two first nodes is internally implemented. For ease of explanation, the embodiments of the present application are illustrated by taking the first node and the first network function as being respectively arranged in different physical devices, or the at least two first nodes involved in the interaction as being respectively arranged in different physical devices, which does not constitute a specific limitation herein.

[0087] In related technologies, AI model parameters can be updated or AI models can be switched in the following ways:

[0088] Method 1: AI model parameter update

[0089] AI model parameter update is to update the AI ​​model parameter values ​​so that the AI ​​model can better match the wireless environment scenario or physical entity configuration changes. It is one of the methods currently discussed in 3GPP R18. Since there is no need to change the AI ​​model structure or AI model deployment, it is an easier AI model evolution method to implement.

[0090] Method 2: AI model switching

[0091] AI model switching refers to a node storing multiple AI models, which are activated and deactivated to match the activated AI model with the changing requirements of the wireless environment scenario or physical entity configuration. It is one of the methods currently being discussed in 3GPP R18. Since the model structure and parameters are fixed for each AI model, and only the switching of the AI ​​model on a fixed node is involved, and the functions implemented by the AI ​​model before and after the switch on the same node remain unchanged, it is an easier AI unit evolution method to implement.

[0092] As can be seen from the above, the AI ​​unit self-evolution method in the relevant technology is only applicable to the scenario where the deployment of the AI ​​unit has not changed, and the evolved AI unit can be used to implement the AI ​​function. However, in the scenario where the AI ​​unit deployment node changes, such as evolving traditional non-AI functions into AI units, or evolving multiple AI units into a single AI unit, the node where the AI ​​unit is deployed does not know how to obtain input data, how to use the input data for model reasoning, and how to apply the model reasoning results, etc., resulting in the inability of the changed AI unit to implement the AI ​​function.

[0093] In the embodiment of the present application, during the evolution of the AI ​​unit, the deployment node and model structure of the evolved first AI unit are not restricted. In this way, the flexibility of the AI ​​unit evolution can be improved, or the interaction process of the evolved AI unit with input data or output data can be reduced. For example, assuming that a certain AI unit is deployed at node A and needs to obtain input data from node B, then through the AI ​​unit evolution, a first AI unit deployed at node B can be obtained. In this way, the first AI unit can obtain input data at the deployment node.

[0094] In addition, in an embodiment of the present application, by sending the first configuration information to the nodes related to the evolved first AI unit, indicating at least one of the node update information, data interaction update information, input data timing change information and output data timing change information related to the first AI unit, in this way, in the scenario where the deployment of the evolved AI unit or the data interaction between the nodes changes, the first configuration information can also be used to enable the nodes related to the first AI unit to collaboratively implement the data collection, model reasoning, application of model reasoning results and other processes of the first AI unit according to the first configuration information, thereby realizing the function of the first AI unit.

[0095] The following, in combination with the accompanying drawings, describes in detail the AI ​​unit configuration method, device and communication equipment provided in the embodiments of the present application through some embodiments and their application scenarios.

[0096] See also Figure 2 , an AI unit configuration method provided in an embodiment of the present application, wherein the execution subject is a first network function, such as Figure 2 As shown, the AI ​​unit configuration method executed by the first network function includes the following steps:

[0097] Step 201: A first network function sends first configuration information to a first node;

[0098] The first node includes a node related to a first AI unit; the first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to the first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on a second node, where N and M are positive integers, and N is not equal to M.

[0099] In some implementations, the M AI units generated based on the N second AI units deployed on the second node may be generated by the N second AI units deployed on the second node through model evolution.

[0100] In some embodiments, the model evolution technology in the embodiments of the present application includes: function fusion or model fusion technology, for example: fusing at least two AI units or fusing at least two functions based on AI unit self-evolution technology to obtain a first AI unit.

[0101] In some other implementations, the model evolution technology in the embodiments of the present application includes: model splitting technology, for example: splitting an AI unit based on the AI ​​unit self-evolution technology to obtain at least two first AI units.

[0102] The embodiments of the present application mainly include the following two application scenarios:

[0103] Scenario 1: The first AI unit includes an AI unit generated based on a non-AI function, which may be a scenario in which a traditional non-AI unit is evolved into an AI unit during a self-evolution process.

[0104] In this application scenario, the nodes related to the first AI unit may include at least one of the following:

[0105] at least one third node on which the first AI unit is deployed;

[0106] A fifth node that provides input data to the first AI unit, such as Figure 3b Node B in

[0107] A fourth node for obtaining the model reasoning result of the first AI unit.

[0108] For example: Figure 3bAs shown, assuming that the first AI unit is deployed at node A, and node C is used to receive the model reasoning result of the first AI unit, the first network function can send the first configuration information to node A to instruct node A to correctly obtain input data, learn the timing relationship of input data, obtain the receiving method of model reasoning result node C, etc., so that node A can correctly implement the input data collection, model reasoning and transmission of model reasoning result of the first AI unit according to the received first configuration information. In addition, the first network function can send the first configuration information to node C, and the first configuration information can indicate the change information related to the source of the model reasoning result that the node C needs to obtain, such as the model reasoning result comes from node A; obtain input data update (for example, compared with non-AI functions, the result predicted by the AI ​​unit is added); obtain the timing change information of the model reasoning result, so as to indicate that node C can obtain accurate model reasoning results from node A and correctly apply the model reasoning results.

[0109] Scenario 2: The first AI unit includes M AI units generated based on N second AI units deployed on the second node, where N and M are positive integers and N is not equal to M. For example, the second AI units deployed on at least two second nodes derive the first AI unit through function fusion or model fusion technology. At this time, at least two old AI units can be fused to obtain a first AI unit. For another example, the second AI unit deployed on a second node obtains the first AI unit deployed on a node through the AI ​​unit self-evolution process, such as Figure 3b The model deployment method shown has evolved into Figure 3a Shown is how the model is deployed.

[0110] The second node represents a node where the old AI unit before fusion is deployed. In some embodiments, two or more old AI units may be fused to obtain a new AI unit, namely the first AI unit. In other embodiments, one or at least two old AI units may be model split to obtain two or more first AI units. For ease of explanation, in the embodiments of the present application, an example of fusing two old AI units to obtain a first AI unit is used for illustration, which is not specifically limited herein.

[0111] In this application scenario, the nodes related to the first AI unit may include at least one of the following:

[0112] at least one third node on which the first AI unit is deployed;

[0113] at least one of the second nodes having the second AI unit deployed thereon;

[0114] A fourth node for obtaining a model reasoning result of the first AI unit;

[0115] A fifth node provides input data to the first AI unit.

[0116] For example: Figure 3a As shown in FIG. 1 , assuming that the number of second AI units is 2, namely, second AI unit 1 deployed at node A and second AI unit 2 deployed at node B, the second node includes node A and node B; thereafter, through self-evolution, the two second AI units are merged into a first AI unit, as shown in FIG. Figure 3c As shown, assuming that the first AI unit is deployed on node B and node C is used to receive the model reasoning result of the first AI unit, the first network function can send respective first configuration information to node A, node B, and node C, so that node A can correctly send part of the input data of the first AI unit to node B according to the received first configuration information; node B can correctly implement the processes of input data collection, model reasoning, and transmission of model reasoning results of the first AI unit according to the received first configuration information; node C can receive the model reasoning result from node B and correctly apply the model reasoning result.

[0117] It should be noted that the second node, the third node, the fourth node and the fifth node may include the same node, for example: Figure 3a As shown, assuming that the number of second AI units is 2, namely, second AI unit 1 deployed at node A and second AI unit 2 deployed at node B, the second node includes node A and node B. Assuming that the deployment of the first AI unit is as follows Figure 3b As shown, node B also serves as the fifth node; node A also serves as the third node.

[0118] In some embodiments, the number of nodes associated with the first AI unit may be one or more, that is, the first node may include one node or at least two nodes, wherein, when the number of first nodes is at least two, the first network function may send respective first configuration information to each first node.

[0119] Optionally, the first configuration information sent by the first network function to different nodes may be the same or different.

[0120] In some embodiments, the first AI unit may be deployed on a third node or at least two third nodes. When the first AI unit is deployed on at least two third nodes, the first AI unit may include at least two AI unit components, each of which is deployed on a corresponding third node.

[0121] For example: Figure 3dAs shown, assuming that the first AI unit includes two AI unit components, one of which is deployed at node A and the other is deployed at node B, the third node includes node A and node B.

[0122] In this implementation manner, the first configuration information may also be used to indicate data interaction information between different third nodes, and timing change information of input data or output data.

[0123] In some implementations, the node update information may explicitly or implicitly indicate any changes to the nodes associated with the first AI unit.

[0124] For example: which input node or nodes have been added, which input node or nodes have been deleted, which output node or nodes have been added, which output node or nodes have been deleted, etc. It can even indicate which node or nodes the first AI unit is deployed on, whether the fourth node used to receive the model inference results of the first AI unit has been updated, etc.

[0125] In this way, the first node can learn which nodes need to interact with data according to the node update information.

[0126] In some implementations, the node update information includes at least one of the following:

[0127] fifth information, the fifth information being used to indicate a predefined node interaction graph;

[0128] The sixth information is used to indicate the corresponding relationship between the sixth node and the nodes in the predefined node interaction graph, wherein the sixth node includes at least one of the following: the first node and a node having a data interaction relationship with the first node.

[0129] The predefined node interaction diagram may be configured by a network-side device or indicated to the first node, or may be agreed upon in a protocol, but is not specifically limited here.

[0130] Optionally, at least two node interaction graphs may be predefined in advance, and each node interaction graph has corresponding identification information. In this case, the fifth information may include the identification information of the node interaction graph.

[0131] For example, the fifth information of 2 bits as shown in Table 1 below may be used to indicate Figure 3a The original deployment shown, and Figure 3b to Figure 3d The three deployment methods of the first AI unit shown (assuming Figure 3b For deployment mode 1, Figure 3c For deployment mode 2, Figure 3c The sequence numbers of the node interaction diagrams corresponding to deployment mode 3) are:

[0132] Table 1

[0133] 00 Figure 3a 01 Figure 3b 02 Figure 3c 03 Figure 3d

[0134] In this implementation, the fifth information can indicate a predefined node interaction graph so that the first node determines the nodes that have a data interaction relationship with the first node, as well as the data interaction mode between the first node and the nodes that have a data interaction relationship, based on the interaction relationship between the nodes in the node interaction graph.

[0135] Optionally, in a predefined node interaction diagram, each node may have its own identification information, such as a node number or a reference number. In this case, the sixth information may indicate the mapping relationship between the first node and the node number or reference number in the node interaction diagram, or the sixth information may also indicate the mapping relationship between a node having a data interaction relationship with the first node and the node number or reference number in the node interaction diagram.

[0136] For example: Assume that the first AI unit is deployed as follows Figure 3b As shown, the predefined node interaction diagram can be indicated by the fifth information, and the predefined node interaction diagram can be indicated by the sixth information. Figure 3b The nodes A, B and C shown in the figure are the corresponding nodes in the node interaction diagram respectively. In this way, the data interaction relationship between the nodes A, B and C can be known according to the node interaction diagram.

[0137] In this implementation, the node update information may implicitly indicate the node update information related to the first AI unit.

[0138] Of course, in some implementations, a default node interaction diagram may be used, in which case it is only necessary to indicate through the sixth information the corresponding node of the first node in the node interaction diagram and the corresponding node of the node that has a data interaction relationship with the first node in the node interaction diagram.

[0139] In some other implementations, the node update information includes at least one of the following:

[0140] first information, the first information including identification information of a deleted input node, the input data of the first AI unit coming from the input node;

[0141] Second information, the second information including identification information of a newly added input node;

[0142] third information, the third information including identification information of the deleted output node, the output data of the first AI unit being used to be transmitted to the output node;

[0143] Fourth information, the fourth information includes identification information of the newly added output node.

[0144] In some embodiments, the input node may include a node for providing input data to the first node; the output node may include a node for obtaining output data of the first node. Figure 3b As shown, for node A, its input nodes include node B, and its output nodes include node C.

[0145] It should be noted that the first configuration information received by different first nodes may be different. For example, assuming that the deployment of the original AI unit is as follows: Figure 3a As shown, through self-evolution, the second AI unit 1 of node A and the second AI unit 2 of node B are merged into the first AI unit. Figure 3b to Figure 3d The three deployment methods of the first AI unit shown (assuming Figure 3b For deployment mode 1, Figure 3c For deployment mode 2, Figure 3c For deployment mode 3), the first indication information in the first configuration information sent by the first network function to node A may be as shown in the following Table 2:

[0146] Table 2

[0147]

[0148]

[0149] The first indication information in the first configuration information sent by the first network function to the node B may be as shown in the following Table 3:

[0150] Table 3

[0151]

[0152] In this implementation manner, the node update information related to the first AI unit may be explicitly indicated by the first indication information.

[0153] 2) The data interaction update information may explicitly or implicitly indicate whether at least one of the input data and output data of the first AI unit is updated, or indicate at least one of the updated input data and output data.

[0154] In this way, after obtaining the data interaction update information, the first node can learn how to obtain input data and how to transmit output data.

[0155] In some implementations, the data interaction update information includes at least one of the following:

[0156] Update information of input data of the first AI unit;

[0157] Update information of output data of the first AI unit.

[0158] In some implementations, the update information of the input data or output data may indicate the data item, data format, data source, etc. of the updated input data or output data, which is not specifically limited herein. The data item may be, for example, a channel impulse response (CIR), a reference signal receiving power (RSRP), a precoding matrix indicator (PMI), etc., which is not specifically limited herein.

[0159] In some implementations, the update information of the input data of the first AI unit includes at least one of the following:

[0160] ninth information, the ninth information being used to indicate input data information of a reference AI unit, wherein the input data of the first AI unit includes the input data of the reference AI unit;

[0161] tenth information, the tenth information is used to indicate a reference data collection pattern, wherein the input data of the first AI unit includes input data collected by the reference data collection pattern;

[0162] The eleventh information is used to indicate a reference AI function, wherein the input data of the first AI unit includes the input data of the reference AI function.

[0163] In some embodiments, the ninth information may include an AI model identifier or an AI function identifier of a reference AI unit, so as to use the input data information of the reference AI unit corresponding to the AI ​​model identifier or the AI ​​function identifier as one of the input data of the first AI unit.

[0164] It should be noted that, taking the example that the ninth information includes the AI ​​model identifier of the reference AI unit, the ninth information may include the AI ​​model identifiers of one or at least two reference AI units, so that the input data information of one or at least two reference AI units may be used as the input data of the first AI unit. For example: the input data of the first node is indicated by referring to the input data of an existing AI unit or multiple existing AI units.

[0165] In some embodiments, the tenth information may be used to refer to a certain data collection mode or multiple data collection modes to indicate the input data of the first node. For example, taking the example that the tenth information includes a reference data set identifier, the tenth information may include one or at least two reference data set identifiers, so that the data collection configuration corresponding to the one or at least two data set identifiers may be used as the collection configuration of the input data of the first AI unit. For example: the collection configuration of the input data of the first node is indicated by referring to an existing data set collection configuration, or multiple existing data set collection configurations.

[0166] In some embodiments, the input data of the first node may be indicated by the eleventh information, with reference to the input of one or at least two AI functions or AI use cases.

[0167] In this implementation, implicit data interaction update information may be used to indicate update information of the input data of the first AI unit.

[0168] In some other embodiments, the update information of the output data of the first AI unit includes at least one of the following:

[0169] Fourteenth information, the fourteenth information is used to indicate output data information of a reference AI unit, wherein the output data of the first AI unit includes the output data of the reference AI unit;

[0170] Fifteenth information, the fifteenth information is used to indicate a reference data collection mode, wherein the output data of the first AI unit includes output data collected in the reference data collection mode;

[0171] Sixteenth information, the sixteenth information is used to indicate a reference AI function, wherein the output data of the first AI unit includes the output data of the reference AI function.

[0172] In some embodiments, the above-mentioned fourteenth information is similar to the ninth information in the previous embodiment, except that the ninth information indicates the input data information of the reference AI unit, and the fourteenth information indicates the output data information of the reference AI unit. For the meaning and function of the fourteenth information, please refer to the explanation of the ninth information in the previous embodiment, which will not be repeated here.

[0173] In some embodiments, the fifteenth information is similar to the tenth information in the previous embodiment, except that the tenth information is used to indicate the input data of the first AI unit, and the fifteenth information is used to indicate the output data of the first AI unit. For the meaning and function of the fifteenth information, reference can be made to the explanation of the tenth information in the previous embodiment, which will not be repeated here.

[0174] In some embodiments, the sixteenth information is similar to the eleventh information in the previous embodiment, except that the eleventh information is used to indicate the input data of the first AI unit, and the sixteenth information is used to indicate the output data of the first AI unit. For the meaning and function of the sixteenth information, reference can be made to the explanation of the eleventh information in the previous embodiment, which will not be repeated here.

[0175] It should be noted that the update information of the input data of the first AI unit and the update information of the output data of the first AI unit may be indicated separately or jointly. For example, when the update information of the input data of the first AI unit and the update information of the output data of the first AI unit are jointly indicated, the ninth information and the fourteenth information may be the same piece of information to indicate that the input data of the reference AI unit is used as one of the input data of the first AI unit, and the output data of the reference AI unit is used as one of the output data of the first AI unit.

[0176] In this implementation manner, at least one of update information of input data of the first AI unit and update information of output data of the first AI unit may be implicitly indicated through data interaction update information.

[0177] In some other embodiments, the update information of the input data of the first AI unit includes at least one of the following:

[0178] Seventh information, the seventh information is used to indicate identification information of newly added input data of the first AI unit;

[0179] The eighth information is used to indicate identification information of input data deleted by the first AI unit.

[0180] In some other embodiments, the update information of the output data of the first AI unit includes at least one of the following:

[0181] Twelfth information, the twelfth information is used to indicate identification information of output data newly added by the first AI unit;

[0182] Thirteenth information, the thirteenth information is used to indicate identification information of output data deleted by the first AI unit.

[0183] In this implementation manner, at least one of update information of input data of the first AI unit and update information of output data of the first AI unit may be explicitly indicated through data interaction update information.

[0184] It should be noted that, in some embodiments, the data interaction update information received by different first nodes may be different. For example, for a certain first node, the data interaction update information received is used to indicate at least one of the update information of the input data and the update information of the output data of the first node.

[0185] Of course, in another embodiment, update information of at least one of the input data and output data of the node related to the first AI unit can be sent to each first node, which is not specifically limited here.

[0186] 3) The timing change information may explicitly or implicitly indicate the time of input data of the first node or the time of output data of the first node.

[0187] In this way, the first node can realize the collection of input data and model reasoning functions of the first AI unit based on the timing of the input data; or realize the transmission of the model reasoning results or model monitoring functions of the first AI unit based on the timing of the output data. For example: the third node deploying the first AI unit collects the input data of the first AI unit in a certain time sequence according to the timing change information of the input data, and uses the first AI unit to perform model reasoning based on the collected input data. Then, the third node outputs the corresponding output data in a specified order according to the timing change information of the output data of the first AI unit, so that the model monitoring function performs model monitoring based on the output data.

[0188] The time of the above-mentioned input data or output data may be indicated by periodic time information, time slot position, symbol position, etc.

[0189] In some implementations, the timing change information is associated with the data interaction update information, and the timing change information is used to indicate at least one of the following:

[0190] Timing change information of input data updated by the first node;

[0191] The timing change information of the output data after the first node is updated.

[0192] In some implementations, the timing change information of the input data can be used to indicate the timing relationship between the input data when the input data of the first node includes at least two, such as which data is obtained first and which data is obtained later. Figure 3b As shown, for the first AI unit deployed on node A, it includes the input data of node A and the input data of node B. At this time, node A can determine when to collect the input data of node A and when to collect the input data of node B based on the timing change information of the input data, so as to realize complete data collection and model reasoning functions.

[0193] In some implementations, the timing change information of the output data may be used to indicate the timing relationship between the respective output data, such as which data is output first and which data is output later, when the output data of the first node includes at least two.

[0194] In some implementations, the timing change information may also be used to indicate the timing relationship between input data and output data of different nodes, for example: Figure 3d As shown, the timing change information is used to instruct node A to output the output data of the second AI unit component of the first AI unit according to the input data, and to instruct node B to input the output data of the second AI unit component and the input data of node B into the first AI unit component in the first AI unit after node A outputs the output data of the second AI unit component in the first AI unit, so as to obtain the output data of node B.

[0195] In this embodiment, the timing change information is associated with the data interaction update information, and can be the update information of at least one of the input data and output data of the first node indicated by the data interaction update information, and then the timing change information of at least one of the updated input data and updated output data of the first node indicated by the timing change information. For example: the first node can first know which input data needs to be obtained or how to obtain the input data according to the data interaction update information, and then the first node can know the acquisition timing of different input data according to the timing change information. In this way, the input data can be input into the first AI unit according to the acquisition timing of different input data, so as to realize the input data collection and model reasoning function of the first AI unit.

[0196] In some implementations, the timing change information includes at least one of the following:

[0197] seventeenth information, the seventeenth information is used to indicate a reference input data item;

[0198] Eighteenth information, the eighteenth information is used to indicate a time deviation between the input data updated by the first node and the reference input data item;

[0199] Nineteenth information, the nineteenth information is used to indicate a reference output data item;

[0200] Twentieth information, the twenty-first information is used to indicate a time deviation between the updated output data of the first node and the reference output data item;

[0201] Twenty-first information, where the twenty-first information is used to indicate a reference signal resource;

[0202] Twenty-second information, the twenty-second information is used to indicate a time deviation between the updated input data of the first node and the time domain position of the reference signal resource;

[0203] Twenty-third information, the twenty-third information is used to indicate a time deviation between the updated output data of the first node and the time domain position of the reference signal resource;

[0204] Twenty-fourth information, the twenty-fourth information is used to indicate target downlink control information (Downlink Control Information, DCI);

[0205] Twenty-fifth information, the twenty-fifth information is used to indicate the time deviation between the updated input data of the first node and the transmission time of the target DCI;

[0206] The twenty-sixth information is used to indicate the time deviation between the updated output data of the first node and the transmission time of the target DCI.

[0207] In one embodiment, the reference input data item indicated by the seventeenth information may be, through the seventeenth information, a certain input data is indicated as a reference input data item, so that the input time of other input data can be referenced by the input time of the reference input data item. For example: through the eighteenth information indicating the time difference between each input data item updated by the first node and the input time of the reference input data item, the first node can determine the time to use each updated input data item according to the eighteenth information and the input time of the reference input data item.

[0208] It is worth mentioning that the reference input data item may be one of the input data updated by the first node, or the input data of other AI units possessed by the first node, or when the first AI unit is fused by at least two second AI units, the reference input data item may be the input data of the second AI unit. The reference input data item is not specifically limited herein.

[0209] In another embodiment, the reference output data item indicated by the nineteenth information may be, through the nineteenth information, a certain output data is indicated as a reference output data item, so that the output time of other output data can be referenced by the output time of the reference output data item. For example: through the twenty-second information, the time difference between each output data item updated by the first node and the output time of the reference output data item is indicated, so that the first node can determine the time to output each updated output data item according to the twentieth information and the output time of the reference output data item.

[0210] Similarly, the reference output data item may be one of the updated output data of the first node, or the output data of other AI units of the first node, or when the first AI unit is fused by at least two second AI units, the reference output data item may be the output data of the second AI unit. The reference output data item is not specifically limited herein.

[0211] In some other implementations, the time domain position of the reference signal resource and the time domain position corresponding to the DCI (such as the time when the terminal receives the DCI) can be used as the reference time, and the time difference between each input data or output data updated by the first node and the reference time can be indicated through the timing change information. In this way, the first node can determine the input time of each input data updated by the first node, or determine the output time of each output data updated by the first node based on the above-mentioned reference time and the time difference between each input data or output data updated by the first node and the reference time.

[0212] In this implementation, the timing change information can implicitly indicate the timing of each input data or output data updated by the first node, so that the first node can obtain input data in the correct timing or output data in the correct timing.

[0213] In the embodiment of the present application, during the evolution of the AI ​​unit, the deployment nodes and model structure of the evolved first AI unit are not restricted, and the first configuration information is sent to the nodes related to the first AI unit through the first network function, so that each node related to the first AI unit can obtain at least one of the node update information, data interaction update information, input data timing change information and output data timing change information related to the first AI unit according to the first configuration information received by each node. In the scenario where the deployment of the AI ​​unit or the data interaction between nodes changes due to the generation of the first AI unit based on a non-AI function, or the second AI units deployed on at least two second nodes are merged into the first AI unit through function fusion or model fusion technology, or the second AI unit deployed on one second node is split into at least two first AI units through AI unit self-evolution, the first configuration information of the embodiment of the present application enables the nodes related to the first AI unit to collaboratively implement the data collection, model reasoning, application of model reasoning results, model monitoring and other processes of the first AI unit according to the first configuration information.

[0214] See also Figure 4 , is a flowchart of another AI unit configuration method provided in an embodiment of the present application, and the execution subject of the another AI unit configuration method may be the first node. Figure 4 As shown, the other AI unit configuration method includes the following steps:

[0215] Step 401: A first node receives first configuration information from a first network function, wherein the first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to a first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on a second node, where N and M are positive integers, and N is not equal to M.

[0216] Step 402: The first node performs a first operation according to the first configuration information. The first operation includes at least one of the following:

[0217] Sending updated input data to at least one third node where the first AI unit is deployed;

[0218] Obtain input data of the first AI unit from the updated input node;

[0219] Based on the first AI unit, perform model reasoning on the updated input data to obtain first output data or model reasoning results;

[0220] Sending second output data to at least one node where a first AI unit component is deployed, wherein the second output data includes output data of a second AI unit component, and the first AI unit includes at least one of the first AI unit component and the second AI unit component;

[0221] Sending the model inference result of the first AI unit to the fourth node;

[0222] Obtain the model inference result of the first AI unit from the third node.

[0223] In some embodiments, the present application provides an AI unit configuration method and Figure 2 The AI ​​unit configuration method shown in the figure is the same as that shown in the figure, except that: Figure 2 The AI ​​unit configuration method shown is executed by the first network function. The AI ​​unit configuration method in this embodiment is executed by the first node, so that after the first network function configures the first configuration information for the first node, the first node can collaboratively implement the first AI unit's data collection, model reasoning, application of model reasoning results, model monitoring and other processes according to the first configuration information.

[0224] It should be noted that the first AI unit, the second AI unit, the first node, the first configuration information, the first network function, the node update information related to the first AI unit, the data interaction update information, the input data timing change information and the output data timing change information in the embodiment of the present application are the same as those in the embodiment of the present application. Figure 2 The first AI unit, the second AI unit, the first node, the first configuration information, the first network function, the node update information related to the first AI unit, the data interaction update information, the input data timing change information and the output data timing change information in the method embodiment shown have the same meaning and function, and will not be repeated here.

[0225] In some embodiments, the first node includes at least one of the following:

[0226] at least one third node on which the first AI unit is deployed;

[0227] at least one of the second nodes having the second AI unit deployed thereon;

[0228] A fourth node for obtaining a model reasoning result of the first AI unit;

[0229] A fifth node provides input data to the first AI unit.

[0230] Optionally, the third node is configured to perform at least one of the following according to the first configuration information:

[0231] Obtain input data of the first AI unit from the updated input node;

[0232] Based on the first AI unit, perform model reasoning on the updated input data to obtain first output data or model reasoning results;

[0233] Sending second output data to at least one node where a first AI unit component is deployed, wherein the second output data includes output data of a second AI unit component, and the first AI unit includes at least one of the first AI unit component and the second AI unit component;

[0234] Send the model inference result of the first AI unit to the fourth node.

[0235] For example: Figure 3b In the deployment method of the first AI unit shown, the third node is node A, which is used to obtain part of the input data from the updated input node (i.e., node B), as well as other input data; then perform model reasoning based on the input data to obtain the model reasoning result, and finally send the model reasoning result to the updated output node (i.e., node C).

[0236] In some implementations, the first network function may send first configuration information to the third node and the updated input node respectively, and the updated input node is used to send updated input data to at least one third node where the first AI unit is deployed according to the received first configuration information; at this time, the third node executes the step of obtaining the input data of the first AI unit from the updated input node, including:

[0237] The third node receives input data from the updated input node.

[0238] In some other implementations, the first network function may not send the first configuration information to the updated input node. In this case, the third node performs the step of obtaining the input data of the first AI unit from the updated input node, including:

[0239] The third node acquires the updated input node according to the first configuration information;

[0240] The third node sends a request message to the updated input node to request the updated input node to provide the input data of the first AI unit to the third node;

[0241] The third node receives input data of the first AI unit.

[0242] In some embodiments, the first AI unit may include at least two AI unit components deployed on at least two third nodes. In this case, output results of the AI ​​unit components, namely, second output data, need to be exchanged between different third nodes.

[0243] It should be noted that the first output data may include the output data of any AI unit component of the first AI unit, and the second output data may include the output data of an AI unit component sent by a third node to another third node, so that the output data of an AI unit component of the first AI unit can be used as the input data of another AI unit component of the first AI unit.

[0244] For example: Figure 3dIn the deployment method of the first AI unit shown, the first AI unit includes a first AI unit component and a second AI unit component. At this time, the third node includes a node B on which the first AI unit component is deployed and a node A on which the second AI unit component is deployed. Node A first obtains the input data of node A, and uses the second AI unit component to perform model reasoning based on the input data of node A to obtain the output data of the second AI unit component. At this time, node A sends the output data of the second AI unit component to node B, so that node B uses the first AI unit component to perform model reasoning based on the output data of the second AI unit component and other input data of node B, obtains the model reasoning result of the first AI unit, and sends it to node C.

[0245] In some implementations, in the above process, assuming that the first node is node A, the first operation performed by node A includes:

[0246] Obtaining input data of the first AI unit from the updated input node, that is, obtaining input data of node A from the updated input node;

[0247] Based on the first AI unit, model reasoning is performed on the updated input data to obtain first output data, that is, model reasoning is performed using the second AI unit component based on the input data of node A to obtain output data of the second AI unit component;

[0248] The second output data is sent to at least one node where the first AI unit component is deployed, that is, the output data of the second AI unit component is sent to the node B as the second output data.

[0249] In some other implementations, in the above process, assuming that the first node is a node B, the first operation performed by the node B includes:

[0250] Obtain input data of the first AI unit from the updated input node, that is, obtain second output data from node A, and obtain input data of node B;

[0251] Based on the first AI unit, model reasoning is performed on the updated input data to obtain a model reasoning result, that is, based on the input data of node B and the output data of node A, model reasoning is performed using the first AI unit component to obtain output data of the first AI unit component, and the output data of the first AI unit component may be the model reasoning result of the first AI unit;

[0252] The model inference result of the first AI unit is sent to the fourth node, that is, the output data of the first AI unit component is sent to node C.

[0253] It is worth noting that the second AI unit component represents a component that needs to send output data to other AI unit components. In different scenarios, the same component can be used as at least one of the second AI unit component and the first AI unit component. For example, assuming that the first AI unit includes three AI unit components, namely AI unit component 1 (deployed on node X), AI unit component 2 (deployed on node Y) and AI unit component 3 (deployed on node Z), AI unit component 1 is used to send output data to AI unit component 2 and AI unit component 3, and AI unit component 2 is used to send output data to AI unit component 3, then the following two transmission situations are included:

[0254] 1) AI unit component 1 is used as the second AI unit component, and AI unit component 2 and AI unit component 3 are used as the first AI unit component. At this time, AI unit component 1 is used to send the output data of AI unit component 1 to AI unit component 2 and AI unit component 3 respectively;

[0255] 2) AI unit component 2 serves as the second AI unit component, and AI unit component 3 serves as the first AI unit component. At this time, AI unit component 2 is used to send the output data of AI unit component 2 to AI unit component 3.

[0256] Optionally, the second node is configured to perform the following operations according to the first configuration information:

[0257] Sending updated input data to at least one third node where the first AI unit is deployed.

[0258] For example: Figure 3a , assuming that the number of second AI units is 2, namely, second AI unit 1 deployed on node A and second AI unit 2 deployed on node B, then the second node includes node A and node B; thereafter, through self-evolution, the two second AI units are merged into a first AI unit, such as Figure 3c As shown, assuming that the first AI unit is deployed at node B, node A is used to send the input data of node A to node B, so that node B uses the first AI unit to perform model inference on the input data of node A and other input data of node B to obtain the model inference result.

[0259] Optionally, the fourth node is configured to perform the following operations according to the first configuration information:

[0260] Obtain the model inference result of the first AI unit from the third node.

[0261] For example: Figure 3b As shown, the fourth node is node C, and the third node is node B, then node C is used to obtain the model reasoning result of node B, and execute subsequent processes such as application of the model reasoning result.

[0262] In some implementations, the first network function may send first configuration information to a third node and a fourth node respectively, and the third node is used to send the model reasoning result of the first AI unit to the fourth node according to the received first configuration information; at this time, the fourth node executes the acquisition of the model reasoning result of the first AI unit from the third node, including:

[0263] The model inference result of the first AI unit is received from the third node.

[0264] In some other implementations, the fourth node executes the step of obtaining the model reasoning result of the first AI unit from the third node, including:

[0265] The fourth node acquires, according to the first configuration information, the third node on which the first AI unit is deployed;

[0266] The fourth node sends a request message to the third node to request the third node to provide the model inference result of the first AI unit to the fourth node;

[0267] The fourth node receives the model inference result of the first AI unit.

[0268] Optionally, the node update information includes at least one of the following:

[0269] fifth information, the fifth information being used to indicate a predefined node interaction graph;

[0270] The sixth information is used to indicate the corresponding relationship between the sixth node and the nodes in the predefined node interaction graph, wherein the sixth node includes at least one of the following: the first node and a node having a data interaction relationship with the first node.

[0271] Optionally, the node update information includes at least one of the following:

[0272] first information, the first information including identification information of a deleted input node, the input data of the first AI unit coming from the input node;

[0273] Second information, the second information including identification information of a newly added input node;

[0274] third information, the third information including identification information of the deleted output node, the output data of the first AI unit being used to be transmitted to the output node;

[0275] Fourth information, the fourth information includes identification information of the newly added output node.

[0276] Optionally, the data interaction update information includes at least one of the following:

[0277] Update information of input data of the first AI unit;

[0278] Update information of output data of the first AI unit.

[0279] Optionally, the update information of the input data of the first AI unit includes at least one of the following:

[0280] ninth information, the ninth information being used to indicate input data information of a reference AI unit, wherein the input data of the first AI unit includes the input data of the reference AI unit;

[0281] tenth information, the tenth information is used to indicate a reference data collection mode, wherein the input data of the first AI unit includes input data collected in the reference data collection mode;

[0282] The eleventh information is used to indicate a reference AI function, wherein the input data of the first AI unit includes the input data of the reference AI function.

[0283] Optionally, the update information of the output data of the first AI unit includes at least one of the following:

[0284] Fourteenth information, the fourteenth information is used to indicate output data information of a reference AI unit, wherein the output data of the first AI unit includes the output data of the reference AI unit;

[0285] Fifteenth information, the fifteenth information is used to indicate a reference data collection mode, wherein the output data of the first AI unit includes output data collected in the reference data collection mode;

[0286] Sixteenth information, the sixteenth information is used to indicate a reference AI function, wherein the output data of the first AI unit includes the output data of the reference AI function.

[0287] Optionally, the update information of the input data of the first AI unit includes at least one of the following:

[0288] Seventh information, the seventh information is used to indicate identification information of newly added input data of the first AI unit;

[0289] The eighth information is used to indicate identification information of input data deleted by the first AI unit.

[0290] Optionally, the update information of the output data of the first AI unit includes at least one of the following:

[0291] Twelfth information, the twelfth information is used to indicate identification information of output data newly added by the first AI unit;

[0292] Thirteenth information, the thirteenth information is used to indicate identification information of output data deleted by the first AI unit.

[0293] Optionally, the timing change information is associated with the data interaction update information, and the timing change information is used to indicate at least one of the following:

[0294] Timing change information of input data updated by the first node;

[0295] The timing change information of the output data after the first node is updated.

[0296] Optionally, the timing change information includes at least one of the following:

[0297] seventeenth information, the seventeenth information is used to indicate a reference input data item;

[0298] Eighteenth information, the eighteenth information is used to indicate a time deviation between the input data updated by the first node and the reference input data item;

[0299] Nineteenth information, the nineteenth information is used to indicate a reference output data item;

[0300] Twentieth information, the twenty-first information is used to indicate a time deviation between the updated output data of the first node and the reference output data item;

[0301] Twenty-first information, where the twenty-first information is used to indicate a reference signal resource;

[0302] Twenty-second information, the twenty-second information is used to indicate a time deviation between the updated input data of the first node and the time domain position of the reference signal resource;

[0303] Twenty-third information, the twenty-third information is used to indicate a time deviation between the updated output data of the first node and the time domain position of the reference signal resource;

[0304] Twenty-fourth information, the twenty-fourth information is used to indicate target downlink control information DCI;

[0305] Twenty-fifth information, the twenty-fifth information is used to indicate the time deviation between the updated input data of the first node and the transmission time of the target DCI;

[0306] The twenty-sixth information is used to indicate the time deviation between the updated output data of the first node and the transmission time of the target DCI.

[0307] In an embodiment of the present application, the nodes related to the first AI unit are respectively used to receive the first configuration information from the first network function, so that each node related to the first AI unit can obtain at least one of the node update information, data interaction update information, input data timing change information and output data timing change information related to the first AI unit according to the first configuration information received by each node. In the scenario where the deployment of the AI ​​unit or the data interaction between nodes changes due to the generation of the first AI unit based on a non-AI function, or the second AI units deployed on at least two second nodes are fused into the first AI unit through function fusion or model fusion technology, or the second AI unit deployed on one second node is split into at least two first AI units through AI unit self-evolution, the nodes related to the first AI unit can collaboratively implement the data collection, model reasoning, application of model reasoning results, model monitoring and other processes of the first AI unit according to the first configuration information.

[0308] The AI ​​unit configuration method provided in the embodiment of the present application can be executed by an AI unit configuration device. In the embodiment of the present application, the AI ​​unit configuration device executing the AI ​​unit configuration method is taken as an example to illustrate the AI ​​unit configuration device provided in the embodiment of the present application.

[0309] like Figure 5 As shown, an AI unit configuration device provided in an embodiment of the present application may be a device corresponding to the first network function, such as Figure 5 As shown, the AI ​​unit configuration device 500 includes the following modules:

[0310] A sending module 501, configured to send first configuration information to a first node;

[0311] The first node includes a node related to a first AI unit; the first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to the first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on a second node, where N and M are positive integers, and N is not equal to M.

[0312] Optionally, the node related to the first AI unit includes at least one of the following:

[0313] at least one third node on which the first AI unit is deployed;

[0314] at least one of the second nodes having the second AI unit deployed thereon;

[0315] A fourth node for obtaining a model reasoning result of the first AI unit;

[0316] A fifth node provides input data to the first AI unit.

[0317] Optionally, the sending module 501 is specifically configured to:

[0318] Send respective first configuration information to each first node.

[0319] Optionally, the first configuration information includes at least one of the following:

[0320] Node update information, where the node update information is used to indicate node update information related to the first AI unit;

[0321] Data interaction update information, where the data interaction update information is used to indicate data interaction update information of the first AI unit;

[0322] Timing change information, where the timing change information is used to indicate timing change information of at least one of input data and output data of the first node.

[0323] Optionally, the node update information includes at least one of the following:

[0324] fifth information, the fifth information being used to indicate a predefined node interaction graph;

[0325] The sixth information is used to indicate the corresponding relationship between the sixth node and the nodes in the predefined node interaction graph, wherein the sixth node includes at least one of the following: the first node and a node having a data interaction relationship with the first node.

[0326] Optionally, the node update information includes at least one of the following:

[0327] first information, the first information including identification information of a deleted input node, the input data of the first AI unit coming from the input node;

[0328] Second information, the second information including identification information of a newly added input node;

[0329] third information, the third information including identification information of the deleted output node, the output data of the first AI unit being used to be transmitted to the output node;

[0330] Fourth information, the fourth information includes identification information of the newly added output node.

[0331] Optionally, the data interaction update information includes at least one of the following:

[0332] Update information of input data of the first AI unit;

[0333] Update information of output data of the first AI unit.

[0334] Optionally, the update information of the input data of the first AI unit includes at least one of the following:

[0335] ninth information, the ninth information being used to indicate input data information of a reference AI unit, wherein the input data of the first AI unit includes the input data of the reference AI unit;

[0336] tenth information, the tenth information is used to indicate a reference data collection mode, wherein the input data of the first AI unit includes input data collected in the reference data collection mode;

[0337] The eleventh information is used to indicate a reference AI function, wherein the input data of the first AI unit includes the input data of the reference AI function.

[0338] 9. The method according to claim 7 or 8, characterized in that the update information of the output data of the first AI unit includes at least one of the following:

[0339] Fourteenth information, the fourteenth information is used to indicate output data information of a reference AI unit, wherein the output data of the first AI unit includes the output data of the reference AI unit;

[0340] Fifteenth information, the fifteenth information is used to indicate a reference data collection mode, wherein the output data of the first AI unit includes output data collected in the reference data collection mode;

[0341] Sixteenth information, the sixteenth information is used to indicate a reference AI function, wherein the output data of the first AI unit includes the output data of the reference AI function.

[0342] Optionally, the update information of the input data of the first AI unit includes at least one of the following:

[0343] Seventh information, the seventh information is used to indicate identification information of newly added input data of the first AI unit;

[0344] The eighth information is used to indicate identification information of input data deleted by the first AI unit.

[0345] Optionally, the update information of the output data of the first AI unit includes at least one of the following:

[0346] Twelfth information, the twelfth information is used to indicate identification information of output data newly added by the first AI unit;

[0347] Twelfth information, the twelfth information is used to indicate identification information of output data deleted by the first AI unit.

[0348] Optionally, the timing change information is associated with the data interaction update information, and the timing change information is used to indicate at least one of the following:

[0349] Timing change information of input data updated by the first node;

[0350] The timing change information of the output data after the first node is updated.

[0351] Optionally, the timing change information includes at least one of the following:

[0352] seventeenth information, the seventeenth information is used to indicate a reference input data item;

[0353] Eighteenth information, the eighteenth information is used to indicate a time deviation between the input data updated by the first node and the reference input data item;

[0354] Nineteenth information, the nineteenth information is used to indicate a reference output data item;

[0355] Twentieth information, the twenty-first information is used to indicate a time deviation between the updated output data of the first node and the reference output data item;

[0356] Twenty-first information, where the twenty-first information is used to indicate a reference signal resource;

[0357] Twenty-second information, the twenty-second information is used to indicate a time deviation between the updated input data of the first node and the time domain position of the reference signal resource;

[0358] Twenty-third information, the twenty-third information is used to indicate a time deviation between the updated output data of the first node and the time domain position of the reference signal resource;

[0359] Twenty-fourth information, the twenty-fourth information is used to indicate target downlink control information DCI;

[0360] Twenty-fifth information, the twenty-fifth information is used to indicate the time deviation between the updated input data of the first node and the transmission time of the target DCI;

[0361] The twenty-sixth information is used to indicate the time deviation between the updated output data of the first node and the transmission time of the target DCI.

[0362] The AI ​​unit configuration device 500 in the embodiment of the present application can be an electronic device, such as an electronic device with an operating system, or a component in an electronic device, such as an integrated circuit or a chip. The electronic device can be a network side device, and the network side device has a first network function.

[0363] The AI ​​unit configuration device 500 provided in the embodiment of the present application can achieve the following Figure 2 The various processes implemented by the illustrated method embodiment achieve the same technical effect and will not be described again here to avoid repetition.

[0364] like Figure 6 As shown, another AI unit configuration device provided in an embodiment of the present application may be a device corresponding to the first node, such as Figure 6 As shown, the AI ​​unit configuration device 600 includes the following modules:

[0365] A receiving module 601 is configured to receive first configuration information from a first network function, wherein the first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to a first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on a second node, where N and M are positive integers, and N is not equal to M;

[0366] The execution module 602 is configured to execute a first operation according to the first configuration information, where the first operation includes at least one of the following:

[0367] Sending updated input data to at least one third node where the first AI unit is deployed;

[0368] Obtain input data of the first AI unit from the updated input node;

[0369] Based on the first AI unit, perform model reasoning on the updated input data to obtain first output data or model reasoning results;

[0370] Sending second output data to at least one node where a first AI unit component is deployed, wherein the second output data includes output data of a second AI unit component, and the first AI unit includes at least one of the first AI unit component and the second AI unit component;

[0371] Sending the model inference result of the first AI unit to the fourth node;

[0372] Obtain the model inference result of the first AI unit from the third node.

[0373] Optionally, the first node includes at least one of the following:

[0374] at least one third node on which the first AI unit is deployed;

[0375] at least one of the second nodes having the second AI unit deployed thereon;

[0376] A fourth node for obtaining a model reasoning result of the first AI unit;

[0377] A fifth node provides input data to the first AI unit.

[0378] Optionally, the node update information includes at least one of the following:

[0379] fifth information, the fifth information being used to indicate a predefined node interaction graph;

[0380] The sixth information is used to indicate the corresponding relationship between the sixth node and the nodes in the predefined node interaction graph, wherein the sixth node includes at least one of the following: the first node and a node having a data interaction relationship with the first node.

[0381] Optionally, the node update information includes at least one of the following:

[0382] first information, the first information including identification information of a deleted input node, the input data of the first AI unit coming from the input node;

[0383] Second information, the second information including identification information of a newly added input node;

[0384] third information, the third information including identification information of the deleted output node, the output data of the first AI unit being used to be transmitted to the output node;

[0385] Fourth information, the fourth information includes identification information of the newly added output node.

[0386] Optionally, the data interaction update information includes at least one of the following:

[0387] Update information of input data of the first AI unit;

[0388] Update information of output data of the first AI unit.

[0389] Optionally, the update information of the input data of the first AI unit includes at least one of the following:

[0390] ninth information, the ninth information being used to indicate input data information of a reference AI unit, wherein the input data of the first AI unit includes the input data of the reference AI unit;

[0391] tenth information, the tenth information is used to indicate a reference data collection mode, wherein the input data of the first AI unit includes input data collected in the reference data collection mode;

[0392] The eleventh information is used to indicate a reference AI function, wherein the input data of the first AI unit includes the input data of the reference AI function.

[0393] Optionally, the update information of the output data of the first AI unit includes at least one of the following:

[0394] Fourteenth information, the fourteenth information is used to indicate output data information of a reference AI unit, wherein the output data of the first AI unit includes the output data of the reference AI unit;

[0395] Fifteenth information, the fifteenth information is used to indicate a reference data collection mode, wherein the output data of the first AI unit includes output data collected in the reference data collection mode;

[0396] Sixteenth information, the sixteenth information is used to indicate a reference AI function, wherein the output data of the first AI unit includes the output data of the reference AI function.

[0397] Optionally, the update information of the input data of the first AI unit includes at least one of the following:

[0398] Seventh information, the seventh information is used to indicate identification information of newly added input data of the first AI unit;

[0399] The eighth information is used to indicate identification information of input data deleted by the first AI unit.

[0400] Optionally, the update information of the output data of the first AI unit includes at least one of the following:

[0401] Twelfth information, the twelfth information is used to indicate identification information of output data newly added by the first AI unit;

[0402] Thirteenth information, the thirteenth information is used to indicate identification information of output data deleted by the first AI unit.

[0403] Optionally, the timing change information is associated with the data interaction update information, and the timing change information is used to indicate at least one of the following:

[0404] Timing change information of input data updated by the first node;

[0405] The timing change information of the output data after the first node is updated.

[0406] Optionally, the timing change information includes at least one of the following:

[0407] seventeenth information, the seventeenth information is used to indicate a reference input data item;

[0408] Eighteenth information, the eighteenth information is used to indicate a time deviation between the input data updated by the first node and the reference input data item;

[0409] Nineteenth information, the nineteenth information is used to indicate a reference output data item;

[0410] Twentieth information, the twentyth information is used to indicate a time deviation between the updated output data of the first node and the reference output data item;

[0411] Twenty-first information, where the twenty-first information is used to indicate a reference signal resource;

[0412] Twenty-second information, the twenty-second information is used to indicate a time deviation between the updated input data of the first node and the time domain position of the reference signal resource;

[0413] Twenty-third information, the twenty-third information is used to indicate a time deviation between the updated output data of the first node and the time domain position of the reference signal resource;

[0414] Twenty-fourth information, the twenty-fourth information is used to indicate target downlink control information DCI;

[0415] Twenty-fifth information, the twenty-fifth information is used to indicate the time deviation between the updated input data of the first node and the transmission time of the target DCI;

[0416] The twenty-sixth information is used to indicate the time deviation between the updated output data of the first node and the transmission time of the target DCI.

[0417] The AI ​​unit configuration device 600 in the embodiment of the present application can be an electronic device, such as an electronic device with an operating system, or a component in an electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal or a network side device. Exemplarily, the terminal can include but is not limited to the types of terminals 11 listed above, and other devices can be servers, network attached storage (NAS), etc., which are not specifically limited in the embodiment of the present application.

[0418] The AI ​​unit configuration device 600 provided in the embodiment of the present application can achieve the following Figure 4 The various processes implemented by the illustrated method embodiment achieve the same technical effect and will not be described again here to avoid repetition.

[0419] like Figure 7 As shown, the embodiment of the present application further provides a communication device 700, including a processor 701 and a memory 702, the memory 702 stores a program or instruction that can be run on the processor 701, for example, when the communication device 700 is a first network function, the program or instruction is executed by the processor 701 to implement the following Figure 2 The steps of the method embodiment shown in the figure can achieve the same technical effect. When the communication device 700 is a first node, the program or instruction is executed by the processor 701 to implement the following Figure 4 The various steps of the method embodiment shown can achieve the same technical effect, and will not be described again here to avoid repetition.

[0420] The embodiment of the present application also provides a terminal, including a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run a program or instruction to implement the following Figure 4 The terminal is used as the first node. At this time, the terminal embodiment corresponds to the method embodiment on the first node side. Each implementation process and implementation method of the method embodiment on the first node side can be applied to the terminal embodiment and can achieve the same technical effect. Specifically, Figure 8 A schematic diagram of the hardware structure of a terminal for implementing an embodiment of the present application.

[0421] The terminal 800 includes but is not limited to: a radio frequency unit 801, a network module 802, an audio output unit 803, an input unit 804, a sensor 805, a display unit 806, a user input unit 807, an interface unit 808, a memory 809 and at least some of the components of a processor 810.

[0422] Those skilled in the art will appreciate that the terminal 800 may also include a power source (such as a battery) for supplying power to various components, and the power source may be logically connected to the processor 810 through a power management system, thereby implementing functions such as managing charging, discharging, and power consumption management through the power management system. Figure 8 The terminal structure shown in the figure does not constitute a limitation on the terminal. The terminal may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently, which will not be described in detail here.

[0423] It should be understood that in the embodiment of the present application, the input unit 804 may include a graphics processing unit (GPU) 8041 and a microphone 8042, and the graphics processing unit 8041 processes the image data of a static picture or video obtained by an image capture device (such as a camera) in a video capture mode or an image capture mode. The display unit 806 may include a display panel 8061, and the display panel 8061 may be configured in the form of a liquid crystal display, an organic light emitting diode, etc. The user input unit 807 includes a touch panel 8071 and at least one of other input devices 8072. The touch panel 8071 is also called a touch screen. The touch panel 8071 may include two parts: a touch detection device and a touch controller. Other input devices 8072 may include, but are not limited to, a physical keyboard, function keys (such as volume control keys, switch keys, etc.), a trackball, a mouse, and a joystick, which will not be repeated here.

[0424] In the embodiment of the present application, after receiving downlink data from the network side device, the radio frequency unit 801 can transmit the data to the processor 810 for processing; in addition, the radio frequency unit 801 can send uplink data to the network side device. Generally, the radio frequency unit 801 includes but is not limited to an antenna, an amplifier, a transceiver, a coupler, a low noise amplifier, a duplexer, etc.

[0425] The memory 809 can be used to store software programs or instructions and various data. The memory 809 may mainly include a first storage area for storing programs or instructions and a second storage area for storing data, wherein the first storage area may store an operating system, an application program or instruction required for at least one function (such as a sound playback function, an image playback function, etc.), etc. In addition, the memory 809 may include a volatile memory or a non-volatile memory. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), a static random access memory (SRAM), a dynamic random access memory (DRAM), a synchronous dynamic random access memory (SDRAM), a double data rate synchronous dynamic random access memory (DDRSDRAM), an enhanced synchronous dynamic random access memory (ESDRAM), a synchronous link dynamic random access memory (SLDRAM) and a direct memory bus random access memory (DRRAM). The memory 809 in the embodiment of the present application includes but is not limited to these and any other suitable types of memory.

[0426] The processor 810 may include one or more processing units; optionally, the processor 810 integrates an application processor and a modem processor, wherein the application processor mainly processes operations related to an operating system, a user interface, and application programs, and the modem processor mainly processes wireless communication signals, such as a baseband processor. It is understandable that the modem processor may not be integrated into the processor 810.

[0427] The radio frequency unit 801 is configured to receive first configuration information from a first network function, wherein the first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to a first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on a second node, where N and M are positive integers, and N is not equal to M;

[0428] The processor 810 is configured to perform a first operation according to the first configuration information, where the first operation includes at least one of the following:

[0429] Control the radio frequency unit 801 to send updated input data to at least one third node where the first AI unit is deployed;

[0430] Control the RF unit 801 to obtain input data of the first AI unit from the updated input node;

[0431] Based on the first AI unit, perform model reasoning on the updated input data to obtain first output data or model reasoning results;

[0432] Control the radio frequency unit 801 to send second output data to at least one node where the first AI unit component is deployed, where the second output data includes output data of the second AI unit component, and the first AI unit includes at least one of the first AI unit component and the second AI unit component;

[0433] Control the radio frequency unit 801 to send the model inference result of the first AI unit to the fourth node;

[0434] The control radio frequency unit 801 obtains the model inference result of the first AI unit from the third node.

[0435] Optionally, the first node includes at least one of the following:

[0436] at least one third node on which the first AI unit is deployed;

[0437] at least one of the second nodes having the second AI unit deployed thereon;

[0438] A fourth node for obtaining a model reasoning result of the first AI unit;

[0439] A fifth node provides input data to the first AI unit.

[0440] Optionally, the first configuration information includes at least one of the following:

[0441] Node update information, where the node update information is used to indicate node update information related to the first AI unit;

[0442] Data interaction update information, where the data interaction update information is used to indicate data interaction update information of the first AI unit;

[0443] Timing change information, where the timing change information is used to indicate timing change information of at least one of input data and output data of the first node.

[0444] Optionally, the node update information includes at least one of the following:

[0445] fifth information, the fifth information being used to indicate a predefined node interaction graph;

[0446] The sixth information is used to indicate the corresponding relationship between the sixth node and the nodes in the predefined node interaction graph, wherein the sixth node includes at least one of the following: the first node and a node having a data interaction relationship with the first node.

[0447] Optionally, the node update information includes at least one of the following:

[0448] first information, the first information including identification information of a deleted input node, the input data of the first AI unit coming from the input node;

[0449] Second information, the second information including identification information of a newly added input node;

[0450] third information, the third information including identification information of the deleted output node, the output data of the first AI unit being used to be transmitted to the output node;

[0451] Fourth information, the fourth information includes identification information of the newly added output node.

[0452] Optionally, the data interaction update information includes at least one of the following:

[0453] Update information of input data of the first AI unit;

[0454] Update information of output data of the first AI unit.

[0455] Optionally, the update information of the input data of the first AI unit includes at least one of the following:

[0456] ninth information, the ninth information being used to indicate input data information of a reference AI unit, wherein the input data of the first AI unit includes the input data of the reference AI unit;

[0457] tenth information, the tenth information is used to indicate a reference data collection mode, wherein the input data of the first AI unit includes input data collected in the reference data collection mode;

[0458] The eleventh information is used to indicate a reference AI function, wherein the input data of the first AI unit includes the input data of the reference AI function.

[0459] Optionally, the update information of the output data of the first AI unit includes at least one of the following:

[0460] Fourteenth information, the fourteenth information is used to indicate output data information of a reference AI unit, wherein the output data of the first AI unit includes the output data of the reference AI unit;

[0461] Fifteenth information, the fifteenth information is used to indicate a reference data collection mode, wherein the output data of the first AI unit includes output data collected in the reference data collection mode;

[0462] Sixteenth information, the sixteenth information is used to indicate a reference AI function, wherein the output data of the first AI unit includes the output data of the reference AI function.

[0463] Optionally, the update information of the input data of the first AI unit includes at least one of the following:

[0464] Seventh information, the seventh information is used to indicate identification information of newly added input data of the first AI unit;

[0465] The eighth information is used to indicate identification information of input data deleted by the first AI unit.

[0466] Optionally, the update information of the output data of the first AI unit includes at least one of the following:

[0467] Twelfth information, the twelfth information is used to indicate identification information of output data newly added by the first AI unit;

[0468] Thirteenth information, the thirteenth information is used to indicate identification information of output data deleted by the first AI unit.

[0469] Optionally, the timing change information is associated with the data interaction update information, and the timing change information is used to indicate at least one of the following:

[0470] Timing change information of input data updated by the first node;

[0471] The timing change information of the output data after the first node is updated.

[0472] Optionally, the timing change information includes at least one of the following:

[0473] seventeenth information, the seventeenth information is used to indicate a reference input data item;

[0474] Eighteenth information, the eighteenth information is used to indicate a time deviation between the input data updated by the first node and the reference input data item;

[0475] Nineteenth information, the nineteenth information is used to indicate a reference output data item;

[0476] Twentieth information, the twenty-first information is used to indicate a time deviation between the updated output data of the first node and the reference output data item;

[0477] Twenty-first information, where the twenty-first information is used to indicate a reference signal resource;

[0478] Twenty-second information, the twenty-second information is used to indicate a time deviation between the updated input data of the first node and the time domain position of the reference signal resource;

[0479] Twenty-third information, the twenty-third information is used to indicate a time deviation between the updated output data of the first node and the time domain position of the reference signal resource;

[0480] Twenty-fourth information, the twenty-fourth information is used to indicate target downlink control information DCI;

[0481] Twenty-fifth information, the twenty-fifth information is used to indicate the time deviation between the updated input data of the first node and the transmission time of the target DCI;

[0482] The twenty-sixth information is used to indicate the time deviation between the updated output data of the first node and the transmission time of the target DCI.

[0483] It can be understood that the implementation process of each implementation method mentioned in this embodiment can refer to the relevant description of the AI ​​unit configuration method embodiment on the first node side, and achieve the same or corresponding technical effect. To avoid repetition, it will not be repeated here.

[0484] The embodiment of the present application also provides a network side device, including a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run a program or instruction to implement the following Figure 2 or Figure 4 The steps of the method embodiment shown. This network side device embodiment corresponds to the above network side device method embodiment, and each implementation process and implementation mode of the above method embodiment can be applied to this network side device embodiment and can achieve the same technical effect.

[0485] Specifically, the embodiment of the present application also provides a network side device. Fig. 9As shown, the network side device 900 includes: an antenna 91, a radio frequency device 92, a baseband device 93, a processor 94 and a memory 95. The antenna 91 is connected to the radio frequency device 92. In the uplink direction, the radio frequency device 92 receives information through the antenna 91 and sends the received information to the baseband device 93 for processing. In the downlink direction, the baseband device 93 processes the information to be sent and sends it to the radio frequency device 92. The radio frequency device 92 processes the received information and sends it out through the antenna 91.

[0486] The method executed by the network-side device in the above embodiment may be implemented in the baseband device 93, which includes a baseband processor.

[0487] The baseband device 93 may include, for example, at least one baseband board on which a plurality of chips are arranged. Fig. 9 As shown, one of the chips is, for example, a baseband processor, which is connected to the memory 95 through a bus interface to call the program in the memory 95 to execute the network device operations shown in the above method embodiment.

[0488] The network side device may further include a network interface 96, which is, for example, a Common Public Radio Interface (CPRI).

[0489] Specifically, the network side device 900 of the embodiment of the present application further includes: instructions or programs stored in the memory 95 and executable on the processor 94, and the processor 94 calls the instructions or programs in the memory 95 to execute. Figure 6 The methods executed by the modules shown achieve the same technical effects, and therefore will not be described here in detail to avoid repetition.

[0490] Specifically, the embodiment of the present application also provides a network side device. Fig.10 As shown, the network side device 1000 includes: a processor 1001, a network interface 1002 and a memory 1003. The network interface 1002 is, for example, a Common Public Radio Interface (CPRI).

[0491] Specifically, the network side device 1000 of the embodiment of the present application further includes: an instruction or program stored in the memory 1003 and executable on the processor 1001, and the processor 1001 calls the instruction or program in the memory 1003 to execute Figure 5 The methods executed by the modules shown achieve the same technical effects, and therefore will not be described here in detail to avoid repetition.

[0492] The embodiment of the present application also provides a readable storage medium on which a program or instruction is stored. When the program or instruction is executed by a processor, the following is implemented: Figure 2 or Figure 4 The various processes of the method embodiment shown can achieve the same technical effect, and will not be described again here to avoid repetition.

[0493] The processor is the processor in the terminal described in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory ROM, a random access memory RAM, a magnetic disk or an optical disk. In some examples, the readable storage medium may be a non-transient readable storage medium.

[0494] The present application embodiment further provides a chip, the chip includes a processor and a communication interface, the communication interface is coupled to the processor, the processor is used to run a program or instruction, to implement the following Figure 2 or Figure 4 The various processes of the method embodiment shown can achieve the same technical effect, and will not be described again here to avoid repetition.

[0495] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0496] The present application embodiment further provides a computer program / program product, wherein the computer program / program product is stored in a storage medium, and the computer program / program product is executed by at least one processor to implement the following Figure 2 or Figure 4 The various processes of the method embodiment shown can achieve the same technical effect, and will not be described again here to avoid repetition.

[0497] The embodiment of the present application also provides a wireless communication system, including: a terminal and a network side device, wherein the terminal can be used to perform the following steps: Figure 4 The steps of the method shown in the figure, the network side device can be used to perform the following steps: Figure 2 ,or Figure 2 and Figure 4 The steps of the method are shown.

[0498] It should be noted that, in this article, the terms "comprise", "include" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises one..." does not exclude the presence of other identical elements in the process, method, article or device including the element. In addition, it should be pointed out that the scope of the method and device in the embodiment of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in reverse order according to the functions involved, for example, the described method may be performed in an order different from that described, and various steps may also be added, omitted or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0499] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of a computer software product plus a necessary general hardware platform, and of course, can also be implemented by hardware. The computer software product is stored in a storage medium (such as ROM, RAM, disk, CD, etc.), including several instructions to enable a terminal or a network-side device to execute the methods described in each embodiment of the present application.

[0500] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms of implementation methods without departing from the purpose of the present application and the scope of protection of the claims, and these implementation methods are all within the protection of the present application.

Claims

1. A method for configuring an artificial intelligence (AI) unit, characterized in that: include: The first network function sends first configuration information to the first node; Wherein, the first node includes a node associated with the first AI unit; The first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to the first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on the second node, N and M are positive integers, and N is not equal to M.

2. The method according to claim 1, characterized in that The node related to the first AI unit includes at least one of the following: at least one third node on which the first AI unit is deployed; at least one of the second nodes having the second AI unit deployed thereon; A fourth node for obtaining a model reasoning result of the first AI unit; A fifth node provides input data to the first AI unit.

3. The method according to claim 1 or 2, characterized in that: The first network function sends first configuration information to the first node, including: The first network function sends respective first configuration information to each first node.

4. The method according to any one of claims 1 to 3, characterized in that The node update information includes at least one of the following: first information, the first information including identification information of a deleted input node, the input data of the first AI unit coming from the input node; Second information, the second information including identification information of a newly added input node; third information, the third information including identification information of the deleted output node, the output data of the first AI unit being used to be transmitted to the output node; Fourth information, the fourth information including identification information of the newly added output node; fifth information, the fifth information being used to indicate a predefined node interaction graph; The sixth information is used to indicate the corresponding relationship between the sixth node and the nodes in the predefined node interaction graph, wherein the sixth node includes at least one of the following: the first node and a node having a data interaction relationship with the first node.

5. The method according to any one of claims 1 to 4, characterized in that The data interaction update information includes at least one of the following: Update information of input data of the first AI unit; Update information of output data of the first AI unit.

6. The method according to claim 5, characterized in that The update information of the input data of the first AI unit includes at least one of the following: seventh information, the seventh information being used to indicate identification information of newly added input data of the first AI unit; eighth information, the eighth information being used to indicate identification information of input data deleted by the first AI unit; ninth information, the ninth information being used to indicate input data information of a reference AI unit, wherein the input data of the first AI unit includes the input data of the reference AI unit; tenth information, the tenth information is used to indicate a reference data collection mode, wherein the input data of the first AI unit includes input data collected in the reference data collection mode; The eleventh information is used to indicate a reference AI function, wherein the input data of the first AI unit includes the input data of the reference AI function.

7. The method according to claim 5 or 6, characterized in that: The update information of the output data of the first AI unit includes at least one of the following: Twelfth information, the twelfth information is used to indicate identification information of output data newly added by the first AI unit; Thirteenth information, the thirteenth information is used to indicate identification information of output data deleted by the first AI unit; Fourteenth information, the fourteenth information is used to indicate output data information of a reference AI unit, wherein the output data of the first AI unit includes the output data of the reference AI unit; Fifteenth information, the fifteenth information is used to indicate a reference data collection mode, wherein the output data of the first AI unit includes output data collected in the reference data collection mode; Sixteenth information, the sixteenth information is used to indicate a reference AI function, wherein the output data of the first AI unit includes the output data of the reference AI function.

8. The method according to any one of claims 5 to 7, characterized in that The timing change information is associated with the data interaction update information, and the timing change information is used to indicate at least one of the following: Timing change information of input data updated by the first node; The timing change information of the output data after the first node is updated.

9. The method according to claim 8, characterized in that The timing change information includes at least one of the following: seventeenth information, the seventeenth information is used to indicate a reference input data item; Eighteenth information, the eighteenth information is used to indicate a time deviation between the input data updated by the first node and the reference input data item; Nineteenth information, the nineteenth information is used to indicate a reference output data item; Twentieth information, the twenty-first information is used to indicate a time deviation between the updated output data of the first node and the reference output data item; Twenty-first information, where the twenty-first information is used to indicate a reference signal resource; Twenty-second information, the twenty-second information is used to indicate a time deviation between the updated input data of the first node and the time domain position of the reference signal resource; Twenty-third information, the twenty-third information is used to indicate a time deviation between the updated output data of the first node and the time domain position of the reference signal resource; Twenty-fourth information, the twenty-fourth information is used to indicate target downlink control information DCI; Twenty-fifth information, the twenty-fifth information is used to indicate the time deviation between the updated input data of the first node and the transmission time of the target DCI; The twenty-sixth information is used to indicate the time deviation between the updated output data of the first node and the transmission time of the target DCI.

10. A method for configuring an artificial intelligence (AI) unit, characterized in that: include: The first node receives first configuration information from the first network function, wherein the first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to the first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on the second node, where N and M are positive integers, and N is not equal to M; The first node performs a first operation according to the first configuration information, where the first operation includes at least one of the following: Sending updated input data to at least one third node where the first AI unit is deployed; Obtain input data of the first AI unit from the updated input node; Based on the first AI unit, perform model reasoning on the updated input data to obtain first output data or model reasoning results; Sending second output data to at least one node where a first AI unit component is deployed, wherein the second output data includes output data of a second AI unit component, and the first AI unit includes at least one of the first AI unit component and the second AI unit component; Sending the model inference result of the first AI unit to the fourth node; Obtain the model inference result of the first AI unit from the third node.

11. The method according to claim 10, characterized in that The first node includes at least one of the following: at least one third node on which the first AI unit is deployed; at least one of the second nodes having the second AI unit deployed thereon; A fourth node for obtaining a model reasoning result of the first AI unit; A fifth node provides input data to the first AI unit.

12. The method according to claim 10 or 11, characterized in that: The node update information includes at least one of the following: first information, the first information including identification information of a deleted input node, the input data of the first AI unit coming from the input node; Second information, the second information including identification information of a newly added input node; third information, the third information including identification information of the deleted output node, the output data of the first AI unit being used to be transmitted to the output node; Fourth information, the fourth information including identification information of the newly added output node; fifth information, the fifth information being used to indicate a predefined node interaction graph; The sixth information is used to indicate the corresponding relationship between the sixth node and the nodes in the predefined node interaction graph, wherein the sixth node includes at least one of the following: the first node and a node having a data interaction relationship with the first node.

13. The method according to any one of claims 10 to 12, characterized in that The data interaction update information includes at least one of the following: Update information of input data of the first AI unit; Update information of output data of the first AI unit.

14. The method according to claim 13, characterized in that The update information of the input data of the first AI unit includes at least one of the following: seventh information, the seventh information being used to indicate identification information of newly added input data of the first AI unit; eighth information, the eighth information being used to indicate identification information of input data deleted by the first AI unit; ninth information, the ninth information being used to indicate input data information of a reference AI unit, wherein the input data of the first AI unit includes the input data of the reference AI unit; tenth information, the tenth information is used to indicate a reference data collection mode, wherein the input data of the first AI unit includes input data collected in the reference data collection mode; The eleventh information is used to indicate a reference AI function, wherein the input data of the first AI unit includes the input data of the reference AI function.

15. The method according to claim 13 or 14, characterized in that The update information of the output data of the first AI unit includes at least one of the following: Twelfth information, the twelfth information is used to indicate identification information of output data newly added by the first AI unit; Thirteenth information, the thirteenth information is used to indicate identification information of output data deleted by the first AI unit; Fourteenth information, the fourteenth information is used to indicate output data information of a reference AI unit, wherein the output data of the first AI unit includes the output data of the reference AI unit; Fifteenth information, the fifteenth information is used to indicate a reference data collection mode, wherein the output data of the first AI unit includes output data collected in the reference data collection mode; Sixteenth information, the sixteenth information is used to indicate a reference AI function, wherein the output data of the first AI unit includes the output data of the reference AI function.

16. The method according to any one of claims 13 to 15, characterized in that The timing change information is associated with the data interaction update information, and the timing change information is used to indicate at least one of the following: Timing change information of input data updated by the first node; The timing change information of the output data after the first node is updated.

17. The method according to claim 16, characterized in that The timing change information includes at least one of the following: seventeenth information, the seventeenth information is used to indicate a reference input data item; Eighteenth information, the eighteenth information is used to indicate a time deviation between the input data updated by the first node and the reference input data item; Nineteenth information, the nineteenth information is used to indicate a reference output data item; Twentieth information, the twenty-first information is used to indicate a time deviation between the updated output data of the first node and the reference output data item; Twenty-first information, where the twenty-first information is used to indicate a reference signal resource; Twenty-second information, the twenty-second information is used to indicate a time deviation between the updated input data of the first node and the time domain position of the reference signal resource; Twenty-third information, the twenty-third information is used to indicate a time deviation between the updated output data of the first node and the time domain position of the reference signal resource; Twenty-fourth information, the twenty-fourth information is used to indicate target downlink control information DCI; Twenty-fifth information, the twenty-fifth information is used to indicate the time deviation between the updated input data of the first node and the transmission time of the target DCI; The twenty-sixth information is used to indicate the time deviation between the updated output data of the first node and the transmission time of the target DCI.

18. An artificial intelligence AI unit configuration device, characterized in that: For a first network function, the apparatus comprises: A sending module, used to send first configuration information to the first node; The first node includes a node related to a first AI unit; the first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to the first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on a second node, where N and M are positive integers, and N is not equal to M.

19. An artificial intelligence AI unit configuration device, characterized in that: For a first node, the apparatus comprises: a receiving module, configured to receive first configuration information from a first network function, wherein the first configuration information is used to configure at least one of node update information, data interaction update information, input data timing change information, and output data timing change information related to a first AI unit; the first AI unit includes an AI unit generated based on a non-AI function, or the first AI unit includes M AI units generated based on N second AI units deployed on a second node, where N and M are positive integers, and N is not equal to M; An execution module is configured to execute a first operation according to the first configuration information, where the first operation includes at least one of the following: Sending updated input data to at least one third node where the first AI unit is deployed; Obtain input data of the first AI unit from the updated input node; Based on the first AI unit, perform model reasoning on the updated input data to obtain first output data or model reasoning results; Sending second output data to at least one node where a first AI unit component is deployed, wherein the second output data includes output data of a second AI unit component, and the first AI unit includes at least one of the first AI unit component and the second AI unit component; Sending the model inference result of the first AI unit to the fourth node; Obtain the model inference result of the first AI unit from the third node.

20. A terminal, characterized in that: It includes a processor and a memory, wherein the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the steps of the artificial intelligence AI unit configuration method as described in any one of claims 10 to 17 are implemented.

21. A network side device, characterized in that: It includes a processor and a memory, wherein the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the steps of the artificial intelligence AI unit configuration method as described in any one of claims 1 to 17 are implemented.

22. A readable storage medium, characterized in that: The readable storage medium stores a program or instruction, and when the program or instruction is executed by the processor, the steps of the artificial intelligence AI unit configuration method as described in any one of claims 1 to 17 are implemented.

Images