Artificial intelligence model processing method and related equipment

Abstract

The invention provides an AI model processing method and related equipment, which are used for improving the performance of an AI model. In the method, a first node determines a first AI model; and the first node sends first information, wherein the first information indicates model information of the first AI model and auxiliary information of the first AI model. Compared with a mode that different nodes only interact respective AI models, the first information indicates the model information of the first AI model, and the first information can also indicate the auxiliary information of the first AI model. Therefore, the receiver of the first information can perform AI model processing (such as training, fusion and the like) on the model information of the first AI model based on the auxiliary information of the first AI model, and the performance of the AI model obtained by processing the receiver of the first information based on the first AI model is improved.

Description

Artificial intelligence model processing method and related equipment Technical Field

[0001] The present application relates to the field of communications, and in particular to an artificial intelligence (AI) model processing method and related equipment. Background Art

[0002] Federated learning (FL) is a widely used model for AI data processing. In a FL system, a central node is used to integrate models, resulting in a star-shaped structure that is prone to poor robustness. If a problem occurs in the central node (e.g., an attack), the entire system can be paralyzed.

[0003] Currently, decentralized AI data processing is an improvement to federated learning AI data processing. This model eliminates the need for a central node and improves system robustness. In a decentralized system, each node trains its own AI model using local data and local targets. It then interacts with its neighboring nodes within reach and further processes (e.g., trains, integrates, etc.) the local AI model based on the interacted AI model.

[0004] However, in the above-mentioned decentralized system, since there is no unified scheduling of the central node, how different distributed nodes implement AI model processing after receiving the AI ​​models of other nodes is a technical problem that needs to be solved urgently.

[0005] Summary of the Invention

[0006] This application provides an AI model processing method and related equipment for improving the performance of AI models.

[0007] The first aspect of the present application provides an AI model processing method, which is executed by a first node, or the method is executed by some components in the first node (such as a processor, a chip or a chip system, etc.), or the method can also be implemented by a logic module or software that can realize all or part of the functions of the first node. In the first aspect and its possible implementation, the method is described as being executed by the first node as an example, and the first node can be a terminal device or a network device. In this method, the first node determines a first AI model; the first node sends first information, and the first information indicates model information of the first AI model and auxiliary information of the first AI model.

[0008] Based on the above technical solution, after determining the first AI model, the first node sends first information indicating the model information of the first AI model and auxiliary information of the first AI model. Compared to a method in which different nodes only exchange their respective AI models, because the first information indicates not only the model information of the first AI model but also the auxiliary information of the first AI model, the recipient of the first information can perform AI model processing (e.g., training, fusion, etc.) on the model information of the first AI model based on the auxiliary information of the first AI model, thereby improving the performance of the AI ​​model obtained by the recipient of the first information based on the first AI model.

[0009] In a possible implementation of the first aspect, the auxiliary information of the first AI model includes at least one of the following: type information of the first AI model, identification information of the first node, identification information of a receiving node of the first AI model, version information of the first AI model, time information of generating the first AI model, geographic location information of generating the first AI model, and distribution information of local data of the first node.

[0010] Based on the above technical solution, the auxiliary information of the first AI model indicated by the first information includes at least one of the above items, so that the recipient of the first information can perform AI model processing on the model information of the first AI model based on the above at least one information, thereby improving the performance of the AI ​​model obtained by the recipient of the first information based on the first AI model.

[0011] In a possible implementation manner of the first aspect, the model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, where the transmission resources include time domain resources and / or frequency domain resources.

[0012] Based on the above technical solution, since the amount of data corresponding to the model information of the first AI model and the amount of data corresponding to the auxiliary information of the first AI model are generally different (for example, the former is generally more than the latter), the two can be carried on different transmission resources respectively.

[0013] Optionally, the transmission resource is a preconfigured resource.

[0014] In a possible implementation manner of the first aspect, the first AI model is obtained based on a node type of the first node.

[0015] Based on the above technical solution, the first AI model indicated by the first information sent by the first node can be a model obtained by the first node based on the node type of the first node. In particular, when the first AI model is obtained based at least on the node type of the first node, the first node can execute different AI model processing processes based on different node types, thereby resolving the issue of single node functionality and improving flexibility.

[0016] In a possible implementation of the first aspect, the first AI model is obtained based on the second AI model and the node type of the first node; the second AI model is obtained based on local data; or, the second AI model is obtained based on K pieces of information, each of the K pieces of information indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, where K is a positive integer; or, the second AI model is obtained based on the local data and the K pieces of information.

[0017] Based on the above technical solution, when the first AI model is derived based on at least the second AI model, the first AI model indicated by the first information sent by the first node can be a model that other nodes can understand, facilitating further model processing by other nodes after receiving the first AI model. Furthermore, the second AI model used to derive the first AI model can be implemented using any of the above methods, enhancing the flexibility of the solution.

[0018] In a possible implementation of the first aspect, the node type includes any one of the following: a node type that performs local training based on local data, a node type that performs fusion processing based on AI models of other nodes, and a node type that performs local training based on the local data and performs fusion processing based on AI models of other nodes.

[0019] Based on the above technical solution, the node type used to obtain the first AI model can be implemented by any of the above items to improve the flexibility of the solution implementation.

[0020] In a possible implementation manner of the first aspect, the method further includes: the first node sending indication information indicating a node type of the first node.

[0021] Based on the above technical solution, the first node can also send indication information indicating the node type of the first node, so that other nodes can clarify the node type of the first node based on the indication information, and subsequently interact with the first node based on the node type of the first node.

[0022] In a possible implementation of the first aspect, the method further includes: the first node determining the node type of the first node based on capability information and / or requirement information; or, the first node receiving indication information indicating the node type of the first node.

[0023] Based on the above technical solution, the first node can clarify the node type of the first node based on its own capability information and / or demand information, or it can clarify the node type of the first node based on the instructions of other nodes, so as to improve the flexibility of the solution implementation.

[0024] In a possible implementation of the first aspect, the first AI model is a model understandable by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2.

[0025] It should be understood that the first AI model is a model that can be understood by M nodes in the system where the first node is located. It can be expressed as the first AI model is a model common to the M nodes in the system where the first node is located, the model structure of the first AI model is a model structure that can be understood by the M nodes in the system where the first node is located, the model structure of the first AI model is a model structure common to the M nodes in the system where the first node is located, the first AI model is a public model of the system where the first node is located, or the model structure of the first AI model is a public model structure of the system where the first node is located.

[0026] Based on the above technical solution, the first AI model indicated by the first information sent by the first node is a model that can be understood by M nodes in the system where the first node is located, so that when there are multiple different node types among the M nodes, other nodes can understand the first AI model and further perform model processing after receiving the first information.

[0027] The second aspect of the present application provides an AI model processing method, which is executed by a second node, or the method is executed by some components in the second node (such as a processor, chip or chip system, etc.), or the method can also be implemented by a logic module or software that can realize all or part of the functions of the second node. In the second aspect and its possible implementation, the method is described as being executed by the second node, and the second node can be a terminal device or a network device. In this method, the second node receives N first information from N first nodes, each of the N first information indicates the model information of the first AI model and the auxiliary information of the first AI model, and N is a positive integer; the second node performs model processing based on the N first information to obtain the target AI model.

[0028] Based on the above technical solution, each of the N first information received by the second node indicates the model information of the first AI model and the auxiliary information of the first AI model. Thereafter, the second node performs model processing based on the N first information to obtain the target AI model. Compared to the method in which different nodes only exchange their respective AI models, since each first information indicates not only the model information of the first AI model but also the auxiliary information of the first AI model, the second node can perform AI model processing (e.g., training, fusion, etc.) on the model information of the first AI model based on the auxiliary information of the first AI model, thereby improving the performance of the AI ​​model obtained by the second node based on the processing of the first AI model.

[0029] Optionally, the target AI model is used to complete the AI ​​task of the second node, or the target AI model is a local model of the second node.

[0030] In a possible implementation of the second aspect, the auxiliary information of the first AI model includes at least one of the following: type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information of generating the first AI model, geographic location information of generating the first AI model, and distribution information of local data of the first node.

[0031] Based on the above technical solution, the auxiliary information of the first AI model indicated by the first information includes at least one of the above items, so that the second node can perform AI model processing on the model information of the first AI model based on the above at least one information, thereby improving the performance of the AI ​​model obtained by the second node based on the first AI model.

[0032] In a possible implementation manner of the second aspect, the model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0033] Based on the above technical solution, since the amount of data corresponding to the model information of the first AI model and the amount of data corresponding to the auxiliary information of the first AI model are generally different (for example, the former is generally more than the latter), the two can be carried on different transmission resources respectively.

[0034] Optionally, the transmission resource is a preconfigured resource.

[0035] In a possible implementation of the second aspect, the second node performs model processing based on the N first information to obtain the target AI model, including: the second node performs model processing based on the N first information and the node type of the second node to obtain the target AI model.

[0036] Based on the above technical solution, the target AI model obtained by the second node after model processing can be a model obtained by the second node based at least on the node type of the second node. In this case, when the target AI model is obtained based at least on the node type of the second node, the second node can perform different AI model processing processes based on different node types, thereby resolving the issue of single node functionality and improving flexibility.

[0037] In a possible implementation of the second aspect, the second node performs model processing based on the N first information and the node type of the second node to obtain the target AI model, including: when the node type of the second node is a node type that performs fusion processing based on the AI ​​models of other nodes, the second node performs model fusion on the N first AI models based on the N first information to obtain the target AI model.

[0038] Based on the above technical solution, when the node type of the second node is a node type that performs fusion processing based on the AI ​​models of other nodes, after the second node receives N first information to determine N first AI models, the second node performs model fusion on the N first AI models to obtain the target AI model.

[0039] In a possible implementation of the second aspect, the second node performs model processing based on the N first information and the node type of the second node to obtain the target AI model, including: when the node type of the second node is a node type that performs local training based on the local data and performs fusion processing based on the AI ​​models of other nodes, the second node performs model fusion on the N first AI models and the second AI model based on the N first information to obtain the target AI model, wherein the second AI model is obtained by training based on local data.

[0040] Based on the above technical solution, when the node type of the second node is a node type that performs local training based on the local data and performs fusion processing based on the AI ​​models of other nodes, after the second node receives N first information to determine N first AI models, the second node needs to obtain a second AI model based on local data training, and perform model fusion on the N first AI models and the second AI model to obtain the target AI model.

[0041] In a possible implementation manner of the second aspect, the method further includes: the second node receiving indication information indicating a node type of the first node.

[0042] Based on the above technical solution, the second node can also receive indication information indicating the node type of the first node, so that the second node can clarify the node type of the first node based on the indication information, and subsequently the second node can interact with the first node based on the node type of the first node.

[0043] In a possible implementation manner of the second aspect, the method further includes: the second node sending indication information for indicating a node type of the second node.

[0044] Based on the above technical solution, the second node can also send indication information indicating the node type of the second node, so that other nodes can clarify the node type of the second node based on the indication information, and subsequently interact with the second node based on the node type of the second node.

[0045] In a possible implementation manner of the second aspect, the method further includes: the second node determining the node type of the second node based on capability information and / or requirement information; or, the second node receiving indication information indicating the node type of the second node.

[0046] Based on the above technical solution, the second node can clarify the node type of the second node based on its own capability information and / or demand information, or it can clarify the node type of the second node based on the instructions of other nodes, so as to improve the flexibility of the solution implementation.

[0047] In a possible implementation of the second aspect, the first AI model is a model understandable by M nodes in the system where the second node is located, where M is an integer greater than or equal to 2.

[0048] It should be understood that the first AI model is a model that can be understood by M nodes in the system where the first node is located. It can be expressed as the first AI model is a model common to the M nodes in the system where the first node is located, the model structure of the first AI model is a model structure that can be understood by the M nodes in the system where the first node is located, the model structure of the first AI model is a model structure common to the M nodes in the system where the first node is located, the first AI model is a public model of the system where the first node is located, or the model structure of the first AI model is a public model structure of the system where the first node is located.

[0049] Based on the above technical solution, among the N first information received by the second node, the first AI model indicated by each first information is a model that can be understood by M nodes in the system where the second node is located, so that when there are multiple different node types among the M nodes, each node can understand the first AI model and further perform model processing after receiving the first information.

[0050] The third aspect of the present application provides an AI model processing method, which is executed by a first node, or the method is executed by some components in the first node (such as a processor, a chip or a chip system, etc.), or the method can also be implemented by a logic module or software that can realize all or part of the functions of the first node. In the third aspect and its possible implementation, the method is described as being executed by the first node as an example, and the first node can be a terminal device or a network device. In this method, the first node obtains a local AI model, and the local AI model is used to complete the AI ​​task of the first node; the first node determines a public AI model based on the local AI model, and the public AI model is a model that can be understood by M nodes in the system where the first node is located, and M is an integer greater than or equal to 2; the first node sends a first message, and the first message is used to indicate the model information of the public AI model.

[0051] Based on the above technical solution, the first node obtains a local AI model for completing the AI ​​task of the first node. The first node determines a public AI model based on the local AI model. The first node sends first information indicating model information of the public AI model. The public AI model is a model that can be understood by M nodes in the system where the first node is located. In other words, the public AI model indicated by the first information sent by the first node is a model that can be understood by M nodes in the system where the first node is located, so that when there are multiple different node types among the M nodes, other nodes can understand the public AI model after receiving the first information and further perform model processing.

[0052] It should be understood that the public AI model is a model that can be understood by M nodes in the system where the first node is located. It can be expressed as the public AI model is a model common to the M nodes in the system where the first node is located, the model structure of the public AI model is a model structure that can be understood by M nodes in the system where the first node is located, and the model structure of the public AI model is a model structure common to the M nodes in the system where the first node is located.

[0053] In a possible implementation manner of the third aspect, the first node determining the public AI model based on the local AI model includes: the first node determining the public AI model based on the local AI model and a node type of the first node.

[0054] Based on the above technical solution, the first node can determine the public AI model based on the local AI model and the node type of the first node, so that the first node can execute different AI model processing processes based on different node types, solving the problem of single node function and improving flexibility.

[0055] In a possible implementation of the third aspect, the node type includes any one of the following: a node type that performs local training based on local data, a node type that performs fusion processing based on the AI ​​model of other nodes, and a node type that performs local training based on the local data and performs fusion processing based on the AI ​​model of other nodes.

[0056] Based on the above technical solution, the node type used to obtain the public AI model can be implemented by any of the above items to improve the flexibility of the solution implementation.

[0057] In a possible implementation manner of the third aspect, the method further includes: the first node sending indication information indicating a node type of the first node.

[0058] Based on the above technical solution, the first node can also send indication information indicating the node type of the first node, so that other nodes can clarify the node type of the first node based on the indication information, and subsequently interact with the first node based on the node type of the first node.

[0059] In a possible implementation of the third aspect, the method further includes: the first node determining the node type of the first node based on capability information and / or requirement information; or, the first node receiving indication information indicating the node type of the first node.

[0060] Based on the above technical solution, the first node can clarify the node type of the first node based on its own capability information and / or demand information, or it can clarify the node type of the first node based on the instructions of other nodes, so as to improve the flexibility of the solution implementation.

[0061] In a possible implementation of the third aspect,

[0062] The local AI model is obtained based on local data; or

[0063] The local AI model is obtained based on K pieces of information, each of the K pieces of information indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, where K is a positive integer; or

[0064] The local AI model is obtained based on the local data and the K information.

[0065] Based on the above technical solution, the local AI model used to obtain the public AI model can be implemented as any of the above items to improve the flexibility of the solution implementation.

[0066] In a possible implementation of the third aspect, the first information further indicates auxiliary information of the public AI model.

[0067] Based on the above technical solution, the first information sent by the first node also indicates auxiliary information of the public AI model. Compared to the method in which different nodes only exchange their respective AI models, because the first information indicates not only the model information of the public AI model, but also the auxiliary information of the public AI model, the recipient of the first information can perform AI model processing (such as training, fusion, etc.) on the model information of the public AI model based on the auxiliary information of the public AI model, thereby improving the performance of the AI ​​model obtained by the recipient of the first information based on the public AI model.

[0068] In a possible implementation of the third aspect, the auxiliary information of the public AI model includes at least one of the following: type information of the public AI model, identification information of the first node, identification information of the receiving node of the public AI model, version information of the public AI model, time information of generating the public AI model, geographic location information of generating the public AI model, and distribution information of local data of the first node.

[0069] Based on the above technical solution, the auxiliary information of the public AI model indicated by the first information includes at least one of the above items, so that the recipient of the first information can perform AI model processing on the model information of the public AI model based on the above at least one information, thereby improving the performance of the AI ​​model obtained by the recipient of the first information based on the public AI model.

[0070] In a possible implementation of the third aspect, the model information of the public AI model and the auxiliary information of the public AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0071] Based on the above technical solution, since the amount of data corresponding to the model information of the public AI model and the amount of data corresponding to the auxiliary information of the public AI model are generally different (for example, the former is generally more than the latter), the two can be carried on different transmission resources respectively.

[0072] The fourth aspect of the present application provides an AI model processing method, which is executed by a second node, or the method is executed by some components in the second node (such as a processor, a chip or a chip system, etc.), or the method can also be implemented by a logic module or software that can realize all or part of the functions of the second node. In the fourth aspect and its possible implementation, the method is described as being executed by the second node as an example, and the second node can be a terminal device or a network device. In this method, the second node receives N first information, and each first information in the N first information is used to indicate the model information of a public AI model, and the public AI model is a model that can be understood by M nodes in the system where the second node is located, and M is an integer greater than or equal to 2; the second node updates the local AI model based on the N first information to obtain an updated local AI model, and the local AI model is used to complete the AI ​​task of the second node.

[0073] Based on the above technical solution, each of the N first information received by the second node is used to indicate model information of a public AI model. Thereafter, the second node updates the local AI model based on the N first information to obtain an updated local AI model. The public AI model is a model that can be understood by the M nodes in the system where the second node is located. In other words, the public AI model indicated by the first information received by the second node is a model that can be understood by the M nodes in the system where the second node is located. This allows other nodes to understand the public AI model and further perform model processing after receiving the first information, even if there are multiple different node types among the M nodes.

[0074] It should be understood that the public AI model is a model that can be understood by M nodes in the system where the second node is located. It can be expressed as the public AI model is a model common to M nodes in the system where the first node is located, the model structure of the public AI model is a model structure that can be understood by M nodes in the system where the first node is located, and the model structure of the public AI model is a model structure common to M nodes in the system where the first node is located.

[0075] In a possible implementation of the fourth aspect, the second node updates the local AI model based on the N first information, and the process of obtaining the updated local AI model includes: the second node updates the local AI model based on the N first information and the node type of the second node, to obtain the updated local AI model.

[0076] Based on the above technical solution, the second node can determine the public AI model based on the N first information and the node type of the second node, so that the second node can execute different AI model processing processes based on different node types, solving the problem of single node function and improving flexibility.

[0077] In a possible implementation of the fourth aspect, the node type includes any one of the following: a node type that performs fusion processing based on the AI ​​model of other nodes, and a node type that performs local training based on the local data and performs fusion processing based on the AI ​​model of other nodes.

[0078] Based on the above technical solution, the node type used to obtain the updated local AI model can be implemented by any of the above items to improve the flexibility of the solution implementation.

[0079] In a possible implementation manner of the fourth aspect, the method further includes: the second node sending indication information indicating a node type of the second node.

[0080] Based on the above technical solution, the second node can also send indication information indicating the node type of the second node, so that other nodes can clarify the node type of the second node based on the indication information, and subsequently interact with the second node based on the node type of the second node.

[0081] In a possible implementation of the fourth aspect, the method further includes: the second node determining the node type of the second node based on capability information and / or requirement information; or, the second node receiving indication information indicating the node type of the second node.

[0082] Based on the above technical solution, the second node can clarify the node type of the second node based on its own capability information and / or demand information, or it can clarify the node type of the second node based on the instructions of other nodes, so as to improve the flexibility of the solution implementation.

[0083] In a possible implementation of the fourth aspect, when the node type is a node type that performs fusion processing based on the AI ​​model of other nodes, the local AI model is obtained based on P pieces of information, each of the P pieces of information indicates model information of the AI ​​model of other nodes and auxiliary information of the AI ​​model of other nodes, and P is a positive integer.

[0084] In a possible implementation of the fourth aspect, when the node type is a node type that performs local training based on the local data and performs fusion processing based on the AI ​​models of other nodes, the local AI model is obtained based on the local data, or the local AI model is obtained based on the local data and the P information, each of the P information indicates model information of the AI ​​model of other nodes and auxiliary information of the AI ​​model of other nodes, and P is a positive integer.

[0085] Based on the above technical solution, in the case of different node types, the local AI model can be implemented in different ways as mentioned above to improve the flexibility of the solution implementation.

[0086] In a possible implementation of the fourth aspect, the first information is also used to indicate auxiliary information of the public AI model.

[0087] Based on the above technical solution, each piece of first information received by the second node also indicates auxiliary information of the public AI model. Compared to a method where different nodes only exchange their own AI models, because the first information indicates not only the model information of the public AI model but also the auxiliary information of the public AI model, the second node can update its local AI model based on the auxiliary information of the public AI model to obtain an updated local AI model, thereby improving the performance of the AI ​​model processed by the second node based on the public AI model.

[0088] In a possible implementation of the fourth aspect, the auxiliary information of the public AI model includes at least one of the following: type information of the public AI model, identification information of the first node, identification information of the receiving node of the public AI model, version information of the public AI model, time information of generating the public AI model, geographic location information of generating the public AI model, and distribution information of local data of the first node.

[0089] Based on the above technical solution, the auxiliary information of the public AI model indicated by the first information includes at least one of the above items, so that the second node can perform AI model processing on the model information of the public AI model based on the above at least one information, thereby improving the performance of the AI ​​model obtained by the second node based on the public AI model.

[0090] In a possible implementation of the fourth aspect, the model information of the public AI model and the auxiliary information of the public AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0091] Based on the above technical solution, since the amount of data corresponding to the model information of the public AI model and the amount of data corresponding to the auxiliary information of the public AI model are generally different (for example, the former is generally more than the latter), the two can be carried on different transmission resources respectively.

[0092] The fifth aspect of the present application provides an AI model processing method, which is executed by a first node, or the method is executed by some components in the first node (such as a processor, chip or chip system, etc.), or the method can also be implemented by a logic module or software that can realize all or part of the functions of the first node. In the fifth aspect and its possible implementation, the method is described as being executed by the first node as an example, and the first node can be a terminal device or a network device. In this method, the first node determines the node type of the first node.

[0093] Based on the above technical solution, the first node can determine its node type, and then the first node can perform AI model processing based on the node type. This allows the first node to perform different AI model processing processes based on different node types, solving the problem of single node function and improving flexibility.

[0094] In a possible implementation manner of the fifth aspect, the first node sends indication information indicating a node type of the first node.

[0095] Based on the above technical solution, the first node can also send indication information indicating the node type of the first node, so that other nodes can clarify the node type of the first node based on the indication information, and subsequently interact with the first node based on the node type of the first node.

[0096] In a possible implementation manner of the fifth aspect, the method further includes: the first node sending first information, where the first information is used to indicate model information of a first AI model, wherein the first AI model is obtained based on a node type of the first node.

[0097] Based on the above technical solution, the first AI model indicated by the first information sent by the first node can be a model obtained by the first node based on the node type of the first node. When the first AI model is obtained based at least on the node type of the first node, the first node can perform different AI model processing based on different node types, thereby resolving the issue of single node functionality and improving flexibility.

[0098] In a possible implementation manner of the fifth aspect, the first AI model is obtained based on the second AI model and the node type of the first node.

[0099] Based on the above technical solution, when the first AI model is obtained based on at least the second AI model, the first AI model indicated by the first information sent by the first node can be a model understandable by other nodes, so that other nodes can perform further model processing after receiving the first AI model.

[0100] In a possible implementation of the fifth aspect,

[0101] The second AI model is obtained based on local data; or

[0102] The second AI model is obtained based on K pieces of information, each of the K pieces of information indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, where K is a positive integer; or

[0103] The second AI model is obtained based on the local data and the K information.

[0104] Based on the above technical solution, the second AI model used to obtain the first AI model can be implemented as any of the above items to improve the flexibility of the solution implementation.

[0105] In a possible implementation of the fifth aspect, the node type includes any one of the following: a node type that performs local training based on local data, a node type that performs fusion processing based on the AI ​​model of other nodes, and a node type that performs local training based on the local data and performs fusion processing based on the AI ​​model of other nodes.

[0106] Based on the above technical solution, the node type used to obtain the first AI model can be implemented by any of the above items to improve the flexibility of the solution implementation.

[0107] In a possible implementation of the fifth aspect, the first information is also used to indicate auxiliary information of the first AI model.

[0108] Based on the above technical solution, the first information sent by the first node also indicates auxiliary information of the first AI model. Compared to the method in which different nodes only exchange their respective AI models, because the first information indicates not only the model information of the first AI model but also the auxiliary information of the first AI model, the recipient of the first information can perform AI model processing (e.g., training, fusion, etc.) on the model information of the first AI model based on the auxiliary information of the first AI model, thereby improving the performance of the AI ​​model obtained by the recipient of the first information based on the first AI model.

[0109] In a possible implementation of the fifth aspect, the auxiliary information of the first AI model includes at least one of the following: type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information of generating the first AI model, geographic location information of generating the first AI model, and distribution information of local data of the first node.

[0110] Based on the above technical solution, the auxiliary information of the first AI model indicated by the first information includes at least one of the above items, so that the recipient of the first information can perform AI model processing on the model information of the first AI model based on the above at least one information, thereby improving the performance of the AI ​​model obtained by the recipient of the first information based on the first AI model.

[0111] In a possible implementation of the fifth aspect, the model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0112] Based on the above technical solution, since the amount of data corresponding to the model information of the public AI model and the amount of data corresponding to the auxiliary information of the public AI model are generally different (for example, the former is generally more than the latter), the two can be carried on different transmission resources respectively.

[0113] In a possible implementation of the fifth aspect, the first AI model is a model that can be understood by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2.

[0114] Based on the above technical solution, the first AI model indicated by the first information sent by the first node is a model that can be understood by M nodes in the system where the first node is located, so that when there are multiple different node types among the M nodes, other nodes can understand the first AI model and further perform model processing after receiving the first information.

[0115] In a possible implementation of the fifth aspect, the method further includes: the first node determining the node type of the first node based on capability information and / or requirement information; or, the first node receiving indication information indicating the node type of the first node.

[0116] Based on the above technical solution, the first node can clarify the node type of the first node based on its own capability information and / or demand information, or it can clarify the node type of the first node based on the instructions of other nodes, so as to improve the flexibility of the solution implementation.

[0117] The sixth aspect of the present application provides an AI model processing method, which is executed by a second node, or the method is executed by some components in the second node (such as a processor, chip or chip system, etc.), or the method can also be implemented by a logic module or software that can realize all or part of the functions of the second node. In the sixth aspect and its possible implementation, the method is described as being executed by the second node, and the second node can be a terminal device or a network device. In this method, the second node receives indication information indicating the node type of the first node; and / or the second node receives first information, and the first information is used to indicate model information of the first AI model, wherein the first AI model is obtained based on the node type of the first node.

[0118] Based on the above technical solution, when the second node receives indication information indicating the node type of the first node, the second node can clarify the node type of the first node based on the indication information, and subsequently the second node can interact with the first node based on the node type of the first node.

[0119] Furthermore, when the second node receives first information indicating model information of the first AI model, the first AI model is obtained based on the node type of the first node. When the first AI model is obtained based at least on the node type of the first node, the first node can execute different AI model processing procedures based on different node types, thereby resolving the issue of single node functionality and improving flexibility.

[0120] In a possible implementation manner of the sixth aspect, the first AI model is obtained based on the node type of the first node and the second AI model.

[0121] Based on the above technical solution, when the first AI model is obtained based on at least the second AI model, the first AI model received by the second node can be a model that can be understood by other nodes other than the first node, so that other nodes (such as the second node) can perform further model processing after receiving the target AI model.

[0122] In a possible implementation of the sixth aspect,

[0123] The second AI model is obtained based on local data; or

[0124] The second AI model is obtained based on K pieces of information, each of the K pieces of information indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, where K is a positive integer; or

[0125] The second AI model is obtained based on the local data and the K information.

[0126] Based on the above technical solution, the second AI model used to obtain the first AI model can be implemented as any of the above items to improve the flexibility of the solution implementation.

[0127] In a possible implementation of the sixth aspect, the node type of the first node includes any one of the following: a node type that performs local training based on local data, a node type that performs fusion processing based on AI models of other nodes, and a node type that performs local training based on the local data and performs fusion processing based on AI models of other nodes.

[0128] Based on the above technical solution, the node type used to obtain the first AI model can be implemented by any of the above items to improve the flexibility of the solution implementation.

[0129] Optionally, the node type of the second node includes any one of the following: a node type that performs fusion processing based on the AI ​​model of other nodes, a node type that performs local training based on the local data and performs fusion processing based on the AI ​​model of other nodes.

[0130] In a possible implementation of the sixth aspect, the first information is also used to indicate auxiliary information of the first AI model.

[0131] Based on the above technical solution, the first information received by the second node also indicates auxiliary information of the first AI model. Compared to a method in which different nodes only exchange their respective AI models, because the first information indicates not only the model information of the first AI model but also the auxiliary information of the first AI model, the second node can perform AI model processing (e.g., training, fusion, etc.) on the model information of the first AI model based on the auxiliary information of the first AI model, thereby improving the performance of the AI ​​model obtained by the second node based on the processing of the first AI model.

[0132] In a possible implementation of the sixth aspect, the auxiliary information of the first AI model includes at least one of the following: type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information of generating the first AI model, geographic location information of generating the first AI model, and distribution information of local data of the first node.

[0133] Based on the above technical solution, the auxiliary information of the first AI model indicated by the first information includes at least one of the above items, so that the recipient of the first information can perform AI model processing on the model information of the first AI model based on the above at least one information, thereby improving the performance of the AI ​​model obtained by the recipient of the first information based on the first AI model.

[0134] In a possible implementation of the sixth aspect, the model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0135] Based on the above technical solution, since the amount of data corresponding to the model information of the public AI model and the amount of data corresponding to the auxiliary information of the public AI model are generally different (for example, the former is generally more than the latter), the two can be carried on different transmission resources respectively.

[0136] In a possible implementation of the sixth aspect, the first AI model is a model that can be understood by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2.

[0137] Based on the above technical solution, the first AI model indicated by the first information received by the second node is a model that can be understood by M nodes in the system where the second node is located, so that when there are multiple different node types among the M nodes, other nodes can understand the first AI model and further perform model processing after receiving the first information.

[0138] In a possible implementation of the sixth aspect, the method further includes: the second node determining the node type of the second node based on capability information and / or requirement information; or, the first node receiving indication information indicating the node type of the second node.

[0139] Based on the above technical solution, the second node can clarify the node type of the second node based on its own capability information and / or demand information, or it can clarify the node type of the second node based on the instructions of other nodes, so as to improve the flexibility of the solution implementation.

[0140] In the seventh aspect of the present application, there is provided an AI model processing device, which is a first node, or the device is a partial component in the first node (such as a processor, chip or chip system, etc.), or the device can also be a logic module or software that can implement all or part of the functions of the first node. In the seventh aspect and its possible implementation, the communication device is described as the first node, and the first node can be a terminal device or a network device.

[0141] The device includes a processing unit and a transceiver unit; the processing unit is used to determine a first AI model; the transceiver unit is used to send first information, where the first information indicates model information of the first AI model and auxiliary information of the first AI model.

[0142] In a possible implementation of the seventh aspect, the auxiliary information of the first AI model includes at least one of the following:

[0143] Type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information for generating the first AI model, geographic location information for generating the first AI model, and distribution information of local data of the first node.

[0144] In a possible implementation of the seventh aspect, the model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0145] In a possible implementation of the seventh aspect, the first AI model is obtained based on a node type of the first node.

[0146] In a possible implementation manner of the seventh aspect, the first AI model is obtained based on the second AI model and the node type of the first node.

[0147] In a possible implementation of the seventh aspect, the second AI model is obtained based on local data; or, the second AI model is obtained based on K pieces of information, each of the K pieces of information indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, where K is a positive integer; or, the second AI model is obtained based on the local data and the K pieces of information.

[0148] In a possible implementation of the seventh aspect, the node type includes any one of the following: a node type that performs local training based on local data, a node type that performs fusion processing based on the AI ​​model of other nodes, and a node type that performs local training based on the local data and performs fusion processing based on the AI ​​model of other nodes.

[0149] In a possible implementation manner of the seventh aspect, the transceiver unit is further used to send indication information indicating the node type of the first node.

[0150] In a possible implementation of the seventh aspect, the processing unit is further used to determine the node type of the first node based on capability information and / or requirement information; or, the transceiver unit is further used to receive indication information indicating the node type of the first node.

[0151] In a possible implementation of the seventh aspect, the first AI model is a model that can be understood by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2.

[0152] In the seventh aspect of the embodiment of the present application, the component modules of the communication device can also be used to execute the steps performed in each possible implementation method of the first aspect and achieve corresponding technical effects. For details, please refer to the first aspect and will not be repeated here.

[0153] In an eighth aspect, the present application provides an AI model processing device, which is a second node, or a partial component (such as a processor, chip, or chip system) in the second node, or a logic module or software capable of implementing all or part of the functions of the second node. In the eighth aspect and its possible implementation, the communication device is described as executing as a second node, and the second node can be a terminal device or a network device.

[0154] The device includes a processing unit and a transceiver unit; the transceiver unit is used to receive N first information, each of the N first information indicates model information of a first AI model and auxiliary information of the first AI model, where N is a positive integer; the processing unit is used to perform model processing based on the N first information to obtain a target AI model.

[0155] Optionally, the target AI model is used to complete the AI ​​task of the second node, or the target AI model is a local model of the second node.

[0156] In a possible implementation of the eighth aspect, the auxiliary information of the first AI model includes at least one of the following:

[0157] Type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information for generating the first AI model, geographic location information for generating the first AI model, and distribution information of local data of the first node.

[0158] In a possible implementation of the eighth aspect, the model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0159] In a possible implementation of the eighth aspect, the processing unit is specifically used to perform model processing based on the N first information and the node type of the second node to obtain the target AI model.

[0160] In a possible implementation of the eighth aspect, when the node type of the second node is a node type that performs fusion processing based on the AI ​​models of other nodes, the processing unit is specifically used to perform model fusion on the N first AI models based on the N first information to obtain the target AI model.

[0161] In a possible implementation of the eighth aspect, when the node type of the second node is a node type that performs local training based on the local data and performs fusion processing based on the AI ​​models of other nodes, the processing unit is specifically used to perform model fusion on the N first AI models and the second AI model based on the N first information to obtain the target AI model, wherein the second AI model is obtained by training based on local data.

[0162] In a possible implementation manner of the eighth aspect, the transceiver unit is further used to receive indication information indicating a node type of the first node.

[0163] In a possible implementation manner of the eighth aspect, the transceiver unit is further used to send indication information for indicating the node type of the second node.

[0164] In a possible implementation of the eighth aspect, the processing unit is further used to determine the node type of the second node based on capability information and / or requirement information; or, the transceiver unit is further used to receive indication information indicating the node type of the second node.

[0165] In a possible implementation of the eighth aspect, the first AI model is a model that can be understood by M nodes in the system where the second node is located, where M is an integer greater than or equal to 2.

[0166] In the eighth aspect of the embodiment of the present application, the constituent modules of the communication device can also be used to execute the steps performed in each possible implementation method of the second aspect and achieve corresponding technical effects. For details, please refer to the second aspect and will not be repeated here.

[0167] The ninth aspect of the present application provides an AI model processing device, which is a first node, or the device is a partial component in the first node (such as a processor, chip or chip system, etc.), or the device can also be a logic module or software that can implement all or part of the functions of the first node. In the ninth aspect and its possible implementation, the communication device is described as an example of executing as a first node, and the first node can be a terminal device or a network device.

[0168] The device includes a processing unit and a transceiver unit; the processing unit is used to obtain a local AI model, which is used to complete the AI ​​task of the first node; the processing unit is also used to determine a public AI model based on the local AI model, where the public AI model is a model understandable by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2; the transceiver unit is used to send first information, where the first information is used to indicate model information of the public AI model.

[0169] In a possible implementation manner of the ninth aspect, the processing unit is specifically configured to determine a public AI model based on the local AI model and a node type of the first node.

[0170] In a possible implementation of the ninth aspect, the node type includes any one of the following: a node type that performs local training based on local data, a node type that performs fusion processing based on the AI ​​model of other nodes, and a node type that performs local training based on the local data and performs fusion processing based on the AI ​​model of other nodes.

[0171] In a possible implementation manner of the ninth aspect, the transceiver unit is further used to send indication information indicating the node type of the first node.

[0172] In a possible implementation of the ninth aspect, the processing unit is further used to determine the node type of the first node based on capability information and / or requirement information; or, the transceiver unit is further used to receive indication information indicating the node type of the first node.

[0173] In a possible implementation of the ninth aspect,

[0174] The local AI model is obtained based on local data; or

[0175] The local AI model is obtained based on K pieces of information, each of the K pieces of information indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, where K is a positive integer; or

[0176] The local AI model is obtained based on the local data and the K information.

[0177] In a possible implementation of the ninth aspect, the first information also indicates auxiliary information of the public AI model.

[0178] In a possible implementation of the ninth aspect, the auxiliary information of the public AI model includes at least one of the following:

[0179] Type information of the public AI model, identification information of the first node, identification information of the receiving node of the public AI model, version information of the public AI model, time information of generating the public AI model, geographic location information of generating the public AI model, and distribution information of local data of the public node.

[0180] In a possible implementation of the ninth aspect, the model information of the public AI model and the auxiliary information of the public AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0181] In the ninth aspect of the embodiment of the present application, the constituent modules of the communication device can also be used to execute the steps performed in each possible implementation method of the third aspect and achieve corresponding technical effects. For details, please refer to the third aspect and will not be repeated here.

[0182] In the tenth aspect of the present application, there is provided an AI model processing device, which is a second node, or the device is a partial component in the second node (such as a processor, chip or chip system, etc.), or the device can also be a logic module or software that can implement all or part of the functions of the second node. In the tenth aspect and its possible implementation, the communication device is described as an example of executing as a second node, and the second node can be a terminal device or a network device.

[0183] The device includes a processing unit and a transceiver unit; the transceiver unit is used to receive N first information, each of the N first information is used to indicate model information of a public AI model, and the public AI model is a model that can be understood by M nodes in the system where the second node is located, where M is an integer greater than or equal to 2; the processing unit is used to update the local AI model based on the N first information to obtain an updated local AI model, and the local AI model is used to complete the AI ​​task of the second node.

[0184] In a possible implementation manner of the tenth aspect, the processing unit is specifically used to update the local AI model based on the N first information and the node type of the second node to obtain an updated local AI model.

[0185] In a possible implementation of the tenth aspect, the node type includes any one of the following: a node type that performs fusion processing based on the AI ​​model of other nodes, and a node type that performs local training based on the local data and performs fusion processing based on the AI ​​model of other nodes.

[0186] In a possible implementation manner of the tenth aspect, the transceiver unit is further used to send indication information indicating the node type of the second node.

[0187] In a possible implementation of the tenth aspect, the processing unit is further used to determine the node type of the second node based on capability information and / or requirement information; or, the transceiver unit is further used to receive indication information indicating the node type of the second node.

[0188] In a possible implementation of the tenth aspect, when the node type is a node type that performs fusion processing based on the AI ​​model of other nodes, the local AI model is obtained based on P pieces of information, each of the P pieces of information indicates model information of the AI ​​model of other nodes and auxiliary information of the AI ​​model of other nodes, and P is a positive integer.

[0189] In a possible implementation of the tenth aspect, when the node type is a node type that performs local training based on the local data and performs fusion processing based on the AI ​​models of other nodes, the local AI model is obtained based on the local data, or the local AI model is obtained based on the local data and the P information, each of the P information indicates model information of the AI ​​model of other nodes and auxiliary information of the AI ​​model of other nodes, and P is a positive integer.

[0190] In a possible implementation of the tenth aspect, the first information is also used to indicate auxiliary information of the public AI model.

[0191] In a possible implementation of the tenth aspect, the auxiliary information of the public AI model includes at least one of the following: type information of the public AI model, identification information of the first node, identification information of the receiving node of the public AI model, version information of the public AI model, time information of generating the public AI model, geographic location information of generating the public AI model, and distribution information of local data of the first node.

[0192] In a possible implementation of the tenth aspect, the model information of the public AI model and the auxiliary information of the public AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0193] In the tenth aspect of the embodiment of the present application, the component modules of the communication device can also be used to execute the steps performed in each possible implementation method of the fourth aspect and achieve corresponding technical effects. For details, please refer to the fourth aspect and will not be repeated here.

[0194] In the eleventh aspect of the present application, there is provided an AI model processing device, which is a first node, or the device is a partial component in the first node (such as a processor, chip or chip system, etc.), or the device can also be a logic module or software that can implement all or part of the functions of the first node. In the eleventh aspect and its possible implementation, the communication device is described as the first node, and the first node can be a terminal device or a network device.

[0195] The device includes a processing unit; the processing unit is used by a first node to determine the node type of the first node.

[0196] In a possible implementation manner of the eleventh aspect, the apparatus further includes a transceiver unit configured to send indication information indicating a node type of the first node.

[0197] In a possible implementation of the eleventh aspect, the device further includes a transceiver unit, which is used to send first information, where the first information is used to indicate model information of the first AI model, wherein the first AI model is obtained based on the node type of the first node.

[0198] Optionally, the first AI model is obtained based on the second AI model and the node type of the first node.

[0199] In a possible implementation of the eleventh aspect,

[0200] The second AI model is obtained based on local data; or

[0201] The second AI model is obtained based on K pieces of information, each of the K pieces of information indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, where K is a positive integer; or

[0202] The second AI model is obtained based on the local data and the K information.

[0203] In a possible implementation of the eleventh aspect, the node type includes any one of the following: a node type that performs local training based on local data, a node type that performs fusion processing based on the AI ​​model of other nodes, and a node type that performs local training based on the local data and performs fusion processing based on the AI ​​model of other nodes.

[0204] In a possible implementation of the eleventh aspect, the first information is also used to indicate auxiliary information of the first AI model, wherein the auxiliary information of the first AI model includes at least one of the following: type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information for generating the first AI model, geographic location information for generating the first AI model, and distribution information of local data of the first node.

[0205] In a possible implementation of the eleventh aspect, the model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0206] In a possible implementation of the eleventh aspect, the first AI model is a model that can be understood by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2.

[0207] In a possible implementation of the eleventh aspect, the processing unit is further used to determine the node type of the first node based on capability information and / or requirement information; or, the transceiver unit is further used to receive indication information indicating the node type of the first node.

[0208] In the eleventh aspect of the embodiment of the present application, the constituent modules of the communication device can also be used to execute the steps performed in the various possible implementation methods of the fifth aspect and achieve corresponding technical effects. For details, please refer to the fifth aspect and will not be repeated here.

[0209] The twelfth aspect of the present application provides an AI model processing device, which is a second node, or the device is a partial component in the second node (such as a processor, chip or chip system, etc.), or the device can also be a logic module or software that can implement all or part of the functions of the second node. In the twelfth aspect and its possible implementation, the communication device is described as an example of executing as a second node, and the second node can be a terminal device or a network device.

[0210] The device includes a transceiver unit;

[0211] The transceiver unit is used for the second node to receive indication information indicating the node type of the first node;

[0212] and / or,

[0213] The transceiver unit is used to receive first information, where the first information is used to indicate model information of the first AI model, wherein the first AI model is obtained based on the node type of the first node.

[0214] Optionally, the first AI model is obtained based on the second AI model and the node type of the first node.

[0215] In a possible implementation of the twelfth aspect,

[0216] The second AI model is obtained based on local data; or

[0217] The second AI model is obtained based on K pieces of information, each of the K pieces of information indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, where K is a positive integer; or

[0218] The second AI model is obtained based on the local data and the K information.

[0219] In a possible implementation of the twelfth aspect, the node type includes any one of the following: a node type that performs local training based on local data, a node type that performs fusion processing based on the AI ​​model of other nodes, and a node type that performs local training based on the local data and performs fusion processing based on the AI ​​model of other nodes.

[0220] In a possible implementation of the twelfth aspect, the first information is also used to indicate auxiliary information of the first AI model, wherein the auxiliary information of the first AI model includes at least one of the following: type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information for generating the first AI model, geographic location information for generating the first AI model, and distribution information of local data of the first node.

[0221] In a possible implementation manner of the twelfth aspect, the model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0222] In a possible implementation of the twelfth aspect, the first AI model is a model that can be understood by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2.

[0223] In a possible implementation of the twelfth aspect, the device also includes a processing unit, which is used to determine the node type of the second node based on capability information and / or requirement information; or, the transceiver unit is also used to receive indication information indicating the node type of the second node.

[0224] In the twelfth aspect of the embodiment of the present application, the constituent modules of the communication device can also be used to execute the steps performed in each possible implementation method of the sixth aspect and achieve corresponding technical effects. For details, please refer to the sixth aspect and will not be repeated here.

[0225] A thirteenth aspect of the embodiments of the present application provides a communication device, including at least one processor, wherein the at least one processor is coupled to a memory;

[0226] The memory is used to store programs or instructions;

[0227] The at least one processor is used to execute the program or instruction so that the device implements the method described in the first aspect or any possible implementation of the first aspect, or so that the device implements the method described in the second aspect or any possible implementation of the second aspect, or so that the device implements the method described in the third aspect or any possible implementation of the third aspect, or so that the device implements the method described in the fourth aspect or any possible implementation of the fourth aspect, or so that the device implements the method described in the fifth aspect or any possible implementation of the fifth aspect, or so that the device implements the method described in the sixth aspect or any possible implementation of the sixth aspect.

[0228] A fourteenth aspect of the embodiments of the present application provides a communication device, including at least one logic circuit and an input and output interface;

[0229] The logic circuit is used to execute the method as described in the first aspect or any possible implementation of the first aspect, or, the logic circuit is used to execute the method as described in the second aspect or any possible implementation of the second aspect, or, the logic circuit is used to execute the method as described in the third aspect or any possible implementation of the third aspect, or, the logic circuit is used to execute the method as described in the fourth aspect or any possible implementation of the fourth aspect, or, the logic circuit is used to execute the method as described in the fifth aspect or any possible implementation of the fifth aspect, or, the logic circuit is used to execute the method as described in the sixth aspect or any possible implementation of the sixth aspect.

[0230] A fifteenth aspect of an embodiment of the present application provides a computer-readable storage medium storing one or more computer-executable instructions. When the computer-executable instructions are executed by a processor, the processor executes the method as described in the first aspect or any possible implementation of the first aspect, or the processor executes the method as described in the second aspect or any possible implementation of the second aspect, or the processor executes the method as described in the third aspect or any possible implementation of the third aspect, or the processor executes the method as described in the fourth aspect or any possible implementation of the fourth aspect, or the processor executes the method as described in the fifth aspect or any possible implementation of the fifth aspect, or the processor executes the method as described in the sixth aspect or any possible implementation of the sixth aspect.

[0231] A sixteenth aspect of an embodiment of the present application provides a computer program product (or computer program) storing one or more computers. When the computer program product is executed by the processor, the processor executes the method of the first aspect or any possible implementation of the first aspect, or the processor executes the method of the second aspect or any possible implementation of the second aspect, or the processor executes the method of the third aspect or any possible implementation of the third aspect, or the processor executes the method of the fourth aspect or any possible implementation of the fourth aspect, or the processor executes the method of the fifth aspect or any possible implementation of the fifth aspect, or the processor executes the method of the sixth aspect or any possible implementation of the sixth aspect.

[0232] A seventeenth aspect of an embodiment of the present application provides a chip system, which includes at least one processor, for supporting a communication device to implement the functions involved in the above-mentioned first aspect or any possible implementation of the first aspect, or for supporting a communication device to implement the functions involved in the above-mentioned second aspect or any possible implementation of the second aspect, or for supporting a communication device to implement the functions involved in the above-mentioned third aspect or any possible implementation of the third aspect, or for supporting a communication device to implement the functions involved in the above-mentioned fourth aspect or any possible implementation of the fourth aspect, or for supporting a communication device to implement the functions involved in the above-mentioned fifth aspect or any possible implementation of the fifth aspect, or for supporting a communication device to implement the functions involved in the above-mentioned sixth aspect or any possible implementation of the sixth aspect.

[0233] In one possible design, the chip system may further include a memory for storing program instructions and data necessary for the first communication device. The chip system may be composed of a chip or may include a chip and other discrete components. Optionally, the chip system may further include an interface circuit that provides program instructions and / or data to the at least one processor.

[0234] An eighteenth aspect of an embodiment of the present application provides a communication system, which includes the communication device of the third aspect and the communication device of the fourth aspect, and / or, the communication system includes the communication device of the fifth aspect and the communication device of the sixth aspect, and / or, the communication system includes the communication device of the seventh aspect and the communication device of the eighth aspect, and / or, the communication system includes the communication device of the ninth aspect and the communication device of the tenth aspect, and / or, the communication system includes the communication device of the eleventh aspect and the communication device of the twelfth aspect, and / or, the communication system includes the communication device of the thirteenth aspect, and / or, the communication system includes the communication device of the fourteenth aspect.

[0235] Among them, the technical effects brought about by any design method in the seventh to eighteenth aspects can refer to the technical effects brought about by the different design methods in the above-mentioned first to sixth aspects, and will not be repeated here.

[0236] It can be seen from the above technical solutions that the solution provided by this application has the following beneficial effects:

[0237] In some embodiments, after determining the first AI model, the first node transmits first information indicating model information of the first AI model and auxiliary information of the first AI model. Compared to a method in which different nodes only exchange their respective AI models, because the first information indicates not only the model information of the first AI model but also the auxiliary information of the first AI model, the recipient of the first information can perform AI model processing (e.g., training, fusion, etc.) on the model information of the first AI model based on the auxiliary information of the first AI model, thereby improving the performance of the AI ​​model obtained by the recipient of the first information based on the first AI model.

[0238] In other embodiments, a first node acquires a local AI model for completing an AI task assigned to the first node, determines a public AI model based on the local AI model, and transmits first information indicating model information of the public AI model. The public AI model is understandable by M nodes in the system in which the first node resides. In other words, the public AI model indicated by the first information transmitted by the first node is understandable by the M nodes in the system in which the first node resides. This allows other nodes to understand the first AI model and perform further model processing after receiving the first information, even if there are multiple different node types among the M nodes.

[0239] In other embodiments, the first node may determine the node type of the first node, and then the first node may perform AI model processing based on the node type. This allows the first node to perform different AI model processing based on different node types, thereby resolving the issue of single node functionality and improving flexibility. BRIEF DESCRIPTION OF THE DRAWINGS

[0240] FIG1a is a schematic diagram of a communication system provided by the present application;

[0241] FIG1b is another schematic diagram of the communication system provided by the present application;

[0242] FIG1c is another schematic diagram of the communication system provided by the present application;

[0243] FIG2a is a schematic diagram of the AI ​​processing process involved in this application;

[0244] FIG2 b is another schematic diagram of the AI ​​processing process involved in this application;

[0245] FIG2c is another schematic diagram of the AI ​​processing process involved in this application;

[0246] FIG2 d is another schematic diagram of the AI ​​processing process involved in this application;

[0247] FIG2e is another schematic diagram of the AI ​​processing process involved in this application;

[0248] Figure 3a is a schematic diagram of the AI ​​processing process based on federated learning;

[0249] FIG3 b is a schematic diagram of the AI ​​processing process based on distributed learning;

[0250] FIG4 is an interactive diagram of the AI ​​model processing method provided in this application;

[0251] FIG5a is a schematic diagram of the AI ​​model processing method provided in this application;

[0252] FIG5 b is another schematic diagram of the AI ​​model processing method provided in this application;

[0253] FIG5c is another schematic diagram of the AI ​​model processing method provided in this application;

[0254] Figure 5d is another schematic diagram of the AI ​​model processing method provided in this application;

[0255] FIG6 is another interactive schematic diagram of the AI ​​model processing method provided in this application;

[0256] FIG7 is another interactive schematic diagram of the AI ​​model processing method provided in this application;

[0257] FIG8 is a schematic diagram of a communication device provided by the present application;

[0258] FIG9 is another schematic diagram of a communication device provided by the present application;

[0259] FIG10 is another schematic diagram of the communication device provided by the present application;

[0260] FIG11 is another schematic diagram of the communication device provided in this application. DETAILED DESCRIPTION

[0261] The following will describe the technical solutions in the embodiments of this application in conjunction with the accompanying drawings. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0262] First, some of the terms used in the embodiments of the present application are explained to facilitate understanding by those skilled in the art.

[0263] (1) Terminal device: It can be a wireless terminal device that can receive network device scheduling and instruction information. The wireless terminal device can be a device that provides voice and / or data connectivity to the user, or a handheld device with wireless connection function, or other processing device connected to a wireless modem.

[0264] Terminal devices can communicate with one or more core networks or the Internet via the RAN. Terminal devices can be mobile terminal devices, such as mobile phones (also known as "cellular" phones, mobile phones), computers, and data cards. For example, they can be portable, pocket-sized, handheld, computer-built-in, or vehicle-mounted mobile devices that exchange voice and / or data with the radio access network. Examples include personal communication service (PCS) phones, cordless phones, Session Initiation Protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (PDAs), tablet computers, and computers with wireless transceiver capabilities. Wireless terminal equipment can also be called system, subscriber unit, subscriber station, mobile station, mobile station (MS), remote station, access point (AP), remote terminal equipment (remote terminal), access terminal equipment (access terminal), user terminal equipment (user terminal), user agent, subscriber station (SS), customer premises equipment (CPE), terminal, user equipment (UE), mobile terminal (MT), etc.

[0265] As an example and not a limitation, in the embodiments of the present application, the terminal device may also be a wearable device. Wearable devices may also be referred to as wearable smart devices or smart wearable devices, etc., which are a general term for wearable devices that are intelligently designed and developed using wearable technology for daily wear, such as glasses, gloves, watches, clothing, and shoes. A wearable device is a portable device that is worn directly on the body or integrated into the user's clothes or accessories. Wearable devices are not only hardware devices, but also achieve powerful functions through software support, data interaction, and cloud interaction. Broadly speaking, wearable smart devices include those that are fully functional, large in size, and can achieve complete or partial functions without relying on smartphones, such as smart watches or smart glasses, etc., as well as those that only focus on a certain type of application function and need to be used in conjunction with other devices such as smartphones, such as various smart bracelets, smart helmets, and smart jewelry for vital sign monitoring.

[0266] The terminal may also be a drone, a robot, a terminal in device-to-device (D2D) communication, a terminal in vehicle-to-everything (V2X), a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal in industrial control, a wireless terminal in self-driving, a wireless terminal in remote medical, a wireless terminal in smart grid, a wireless terminal in transportation safety, a wireless terminal in smart city, a wireless terminal in smart home, etc.

[0267] In addition, the terminal device may also be a terminal device in a communication system evolved after the fifth generation (5G) communication system (e.g., a sixth generation (6G) communication system) or a terminal device in a future-evolved public land mobile network (PLMN). For example, a 6G network may further expand the form and function of a 5G communication terminal, and 6G terminals include but are not limited to vehicles, cellular network terminals (with integrated satellite terminal functions), drones, and Internet of Things (IoT) devices.

[0268] In an embodiment of the present application, the terminal device may also obtain AI services provided by the network device. Optionally, the terminal device may also have AI processing capabilities.

[0269] (2) Network equipment: It can be a device in a wireless network. For example, the network equipment can be a RAN node (or device) that connects a terminal device to a wireless network, which can also be called a base station. Currently, some examples of RAN equipment include: base stations gNB (gNodeB) in 5G communication systems, transmission reception points (TRP), evolved Node B (eNB), radio network controllers (RNC), Node B (NB), home base stations (e.g., home evolved Node B, or home Node B, HNB), base band units (BBU), or wireless fidelity (Wi-Fi) access points AP, etc. In addition, in a network structure, the network equipment can include centralized unit (CU) nodes, distributed unit (DU) nodes, or RAN equipment including CU nodes and DU nodes.

[0270] The network device may be any other device that provides wireless communication functionality to the terminal device. The embodiments of this application do not limit the specific technology and device form used by the network device. For ease of description, the embodiments of this application do not limit this.

[0271] The network equipment may also include core network equipment, which may include, for example, a mobility management entity (MME), a home subscriber server (HSS), a serving gateway (S-GW), a policy and charging rules function (PCRF), and a public data network gateway (PDN gateway, P-GW) in a fourth generation (4G) network; and network elements such as an access and mobility management function (AMF), a user plane function (UPF), or a session management function (SMF) in a 5G network. In addition, the core network equipment may also include other core network equipment in a 5G network and a next generation network of a 5G network.

[0272] In an embodiment of the present application, the above-mentioned network device may also have a network node with AI capabilities, which can provide AI services for terminals or other network devices. For example, it can be an AI node on the network side (access network or core network), a computing power node, a RAN node with AI capabilities, a core network element with AI capabilities, etc.

[0273] In the embodiments of the present application, the apparatus for implementing the function of the network device may be the network device, or may be a device capable of supporting the network device in implementing the function, such as a chip system, which may be installed in the network device. In the technical solutions provided in the embodiments of the present application, the technical solutions provided in the embodiments of the present application are described by taking the network device as an example.

[0274] (3) Configuration and pre-configuration: In this application, configuration and pre-configuration are used simultaneously. Configuration refers to the network device / server sending some parameter configuration information or parameter values ​​to the terminal through messages or signaling, so that the terminal can determine the communication parameters or resources during transmission based on these values ​​or information. Pre-configuration is similar to configuration, and can be parameter information or parameter values ​​pre-negotiated between the network device / server and the terminal device, or parameter information or parameter values ​​used by the base station / network device or terminal device as specified in the standard protocol, or parameter information or parameter values ​​pre-stored in the base station / server or terminal device. This application does not limit this.

[0275] Furthermore, these values ​​and parameters can be changed or updated.

[0276] (4) The terms "system" and "network" in the embodiments of the present application can be used interchangeably. "Multiple" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, "at least one of A, B and C" includes A, B, C, AB, AC, BC or ABC. In addition, unless otherwise specified, the ordinal numbers such as "first" and "second" mentioned in the embodiments of the present application are used to distinguish multiple objects, and are not used to limit the order, timing, priority or importance of multiple objects.

[0277] In this application, unless otherwise specified, the same or similar parts between the various embodiments can refer to each other. In the various embodiments of this application, and the various methods / designs / implementations in each embodiment, if there is no special explanation and logical conflict, the terms and / or descriptions between different embodiments and the various methods / designs / implementations in each embodiment are consistent and can be referenced to each other. The technical features in different embodiments and the various methods / designs / implementations in each embodiment can be combined to form new embodiments, methods, or implementations according to their inherent logical relationships. The following description of the implementation methods of this application does not constitute a limitation on the scope of protection of this application.

[0278] The present application can be applied to a long term evolution (LTE) system, a new radio (NR) system, or a communication system evolved after 5G (such as 6G, etc.). The communication system includes at least one network device and / or at least one terminal device.

[0279] Please refer to Figure 1a, which is a schematic diagram of a communication system in this application. Figure 1a exemplarily illustrates a network device 101 and six terminal devices, namely terminal device 1, terminal device 2, terminal device 3, terminal device 4, terminal device 5, and terminal device 6. In the example shown in Figure 1a, terminal device 1 is a smart teacup, terminal device 2 is a smart air conditioner, terminal device 3 is a smart gas pump, terminal device 4 is a vehicle, terminal device 5 is a mobile phone, and terminal device 6 is a printer.

[0280] As shown in Figure 1a, the AI ​​configuration information sending entity can be a network device. The AI ​​configuration information receiving entity can be terminal devices 1-6. In this case, the network device and terminal devices 1-6 form a communication system. In this communication system, terminal devices 1-6 can send data to the network device, and the network device needs to receive data sent by terminal devices 1-6. At the same time, the network device can send configuration information to terminal devices 1-6.

[0281] Optionally, the AI ​​configuration information may include indication information of the node type mentioned later; the data may include model information of the AI ​​model and / or auxiliary information of the AI ​​model mentioned later.

[0282] For example, in Figure 1a, terminal devices 4 and 6 can also form a communication system. Terminal device 5 serves as a network device, i.e., the AI ​​configuration information sending entity; terminal devices 4 and 6 serve as terminal devices, i.e., the AI ​​configuration information receiving entities. For example, in a connected vehicle system, terminal device 5 sends AI configuration information to terminal devices 4 and 6, respectively, and receives data from them. Correspondingly, terminal devices 4 and 6 receive AI configuration information from terminal device 5 and send data to terminal device 5.

[0283] Taking the communication system shown in Figure 1a as an example, in addition to performing communication-related services, different devices (including between network devices and network devices, between network devices and terminal devices, and / or between terminal devices) may also perform AI-related services. For example, as shown in Figure 1b, taking the network device as a base station as an example, the base station can perform communication-related services and AI-related services with one or more terminal devices, and different terminal devices can also perform communication-related services and AI-related services. For another example, as shown in Figure 1c, taking the terminal devices including a TV and a mobile phone as an example, communication-related services and AI-related services can also be performed between the TV and the mobile phone.

[0284] With the advent of the big data era, every device (including network devices and terminal devices) generates massive amounts of raw data in various forms. This data is created as "islands" and exists in every corner of the world. Traditional centralized learning requires each distributed device to transmit local data to a central server, which then uses the collected data for model training and learning. However, this architecture has been increasingly limited by the following factors:

[0285] (1) Distributed devices are widely distributed in various regions and corners of the world. These devices will continuously generate and accumulate huge amounts of raw data at a rapid rate. If the central end needs to collect raw data from all distributed devices, it will inevitably bring huge communication losses and computing power requirements.

[0286] (2) As real-life scenarios become more complex, more and more learning tasks require distributed devices to make timely and effective decisions and provide feedback. Traditional centralized learning involves uploading large amounts of data, which inevitably leads to a large degree of delay, making it unable to meet the real-time needs of actual task scenarios.

[0287] (3) Considering issues such as industry competition, user privacy, and complex administrative procedures, centralized data integration will face increasing resistance and constraints. Therefore, system deployment will increasingly tend to store data locally, while distributed devices will perform local model calculations.

[0288] Therefore, designing a machine learning framework that allows AI systems to more efficiently and accurately share their data while meeting data privacy, security, and regulatory requirements has become a key issue in the current development of artificial intelligence. Machine learning is a key research area in the field of artificial intelligence, and research on neural networks has been a hot topic in recent years. The following is a brief introduction to the neural networks that may be involved in this invention.

[0289] 1. Fully connected neural network.

[0290] Also known as a multilayer perceptron (MLP), an MLP consists of an input layer (left), an output layer (right), and multiple hidden layers (center). Each layer of the MLP contains several nodes, called neurons. Neurons in adjacent layers are connected to each other.

[0291] Alternatively, considering neurons in two adjacent layers, the output h of the neurons in the next layer is the weighted sum of all neurons x in the previous layer connected to it and passes through the activation function, which can be expressed as:

[0292] h=f(wx+b).

[0293] Among them, w is the weight matrix, b is the bias vector, and f is the activation function.

[0294] Alternatively, the output of the neural network can be recursively expressed as:

[0295] y=f n (w n f n-1 (…)+bn ).

[0296] Where n is the index of the neural network layer, 1<=n<=N, where N is the total number of neural network layers.

[0297] In other words, a neural network can be understood as a mapping from an input data set to an output data set. Neural networks are typically initialized randomly, and the process of obtaining this mapping from random w and b using existing data is called neural network training.

[0298] Optionally, the specific training method is to use a loss function to evaluate the output of the neural network. As shown in Figure 2b, the error can be backpropagated, and the neural network parameters (including w and b) can be iteratively optimized through gradient descent until the loss function reaches a minimum value, i.e., the "optimal point" in Figure 2b. It is understandable that the neural network parameters corresponding to the "optimal point" in Figure 2b can be used as the neural network parameters in the trained AI model information.

[0299] Alternatively, the gradient descent process can be expressed as:

[0300]

[0301] Among them, θ is the parameter to be optimized (including w and b), L is the loss function, and η is the learning rate, which controls the step size of gradient descent. represents the derivative operation, represents the derivative of θ with respect to L.

[0302] Alternatively, the backpropagation process can utilize the chain rule for partial derivatives. As shown in Figure 2c, the gradient of the parameters of the previous layer can be recursively calculated from the gradient of the parameters of the next layer, which can be expressed as:

[0303]

[0304] Among them, w ij is the weight of node j connecting to node i, s i is the weighted sum of the inputs to node i.

[0305] 2. Generative Adversarial Network (GAN).

[0306] Generative models (GMs) have been a rapidly developing research area in computer vision in recent years. In 2014, Ian Goodfellow proposed a generative model based on probability distributions for generating realistic data, named the Generative Adversarial Network (GAN). Due to its ability to approximate complex probability density functions, GANs have been proven to be applicable to a variety of machine learning tasks, including image generation, video generation, and natural language processing.

[0307] As shown in Figure 2d, a generative adversarial network typically consists of at least one generator and at least one discriminator. The generator randomly samples from a latent space as input, processes it, and outputs fake samples that are required to closely resemble real samples in the training set. The discriminator takes as input a real sample or a fake sample from the generator and outputs a probability value that the input sample is a real sample. The discriminator aims to distinguish fake samples produced by the generator from real samples, while the generator needs to deceive the discriminator as much as possible, making it unable to distinguish the authenticity of the input sample. By continuously competing with each other and adjusting their respective model parameters, the discriminator is ultimately unable to determine whether the generator's output sample is a real sample.

[0308] 3. Autoencoder.

[0309] An autoencoder is an artificial neural network structure, as shown in Figure 2e. It is generally used to represent (also called encode) a set of data, often for dimensionality reduction. It consists of an encoder and a decoder. The encoder encodes the input data through one or more layers of a neural network, generating a reduced-dimensional codeword. The decoder reconstructs the codeword into output data using one or more layers of a neural network. This output data is required to be as identical as possible to the encoder input data. The encoder and decoder are trained simultaneously to achieve this.

[0310] In recent years, AI technology has made significant progress in areas such as machine vision and natural language processing, and is gradually becoming ubiquitous in real life. It is foreseeable that AI will become ubiquitous in various connected devices (such as terminals and edge devices). In some possible implementations, communication systems can become platforms for large-scale machine learning and AI services. Terminal devices can not only access AI inference services and AI model training services from the network, but also participate in data collection for network model training and even participate in distributed model training.

[0311] In one possible implementation, federated learning (FL) is widely used as a typical AI data processing model. The concept of FL effectively solves the current difficulties faced by the development of artificial intelligence. It fully protects the privacy and security of user data by promoting the collaboration between various edge devices and central servers to efficiently complete the model learning task. As shown in Figure 3a, the FL architecture is the most widely used training architecture in the current FL field. The FedAvg algorithm is the basic algorithm of FL. Taking the central node as the central end and the distributed nodes as the client as an example, the algorithm flow is roughly as follows:

[0312] (1) The center initializes the model to be trained And broadcast it to all client devices.

[0313] (2) In the round t∈[1,T], client k∈[1,K] based on the local dataset For the received global model Perform E epochs of training to obtain local training results Report it to the central node.

[0314] (3) The central node aggregates and collects the local training results from all (or some) clients. Assume that the client set that uploads the local model in round t is The center will use the number of samples of the corresponding client as the weight to perform weighted averaging to obtain a new global model. The specific update rule is: The center then sends the latest version of the global model Broadcast to all client devices for a new round of training.

[0315] (4) Repeat steps (2) and (3) until the model finally converges or the number of training rounds reaches the upper limit.

[0316] Alternatively, convergence generally refers to the performance of the trained model meeting preset requirements. For example, when using FL to train a model for image classification tasks, with a preset classification accuracy of 95%, as training progresses, the model's accuracy performance is evaluated using a test set. When the accuracy reaches 95% or above, the model is considered converged and training is stopped. However, poorly designed factors such as model structure design, parameter selection, and training methods may result in the preset performance requirements being permanently exceeded. In this case, it is necessary to consider setting an upper limit for the number of training rounds. Once the upper limit is reached, training should be stopped even if the model has not achieved the preset performance.

[0317] In addition to reporting local AI models You can also use the local gradient of training After reporting, the central node averages the local gradients and updates the global model according to the direction of the average gradient.

[0318] As you can see, in the FL framework, datasets exist on distributed nodes. Distributed nodes collect local datasets, perform local training, and report the local training results (models or gradients) to the central node. The central node itself does not have a dataset; it is only responsible for fusing the training results of distributed nodes to obtain a global model and send it to the distributed nodes.

[0319] As can be seen from the above description, the federated learning method requires a central node in the system to fuse the models, resulting in a star-shaped structure. This has the problem of poor robustness. Once a problem occurs in the central node (such as being attacked), the entire system will be paralyzed. At the same time, in the federated learning system, the premise for the fusion of multiple local AI models using weighted averaging is that each local AI model has the same structure and parameters. However, in actual networks, the device capabilities of each distributed node are different. Requiring them to use the same local AI model often limits their flexibility and cannot meet the different needs of different distributed nodes.

[0320] Another possible implementation is a decentralized AI data processing model, an improvement over the federated learning AI data processing model. This model eliminates the need for a central node and can improve system robustness. In a decentralized system, each node calculates its own AI model using local data and local targets. It then exchanges its own AI model with reachable neighboring nodes, and further processes (e.g., trains, fuses, etc.) the local AI model based on the exchanged AI model.

[0321] For example, unlike federated learning, the decentralized learning process in the decentralized AI data processing model is shown in Figure 3b, that is, a fully distributed system without a central node. The design goal f(x) of this system is generally the goal f of each node. i The mean of (x), that is Where n is the number of distributed nodes, x is the parameter to be optimized. In machine learning, x is the parameter of the machine learning (such as neural network) model. Each node uses local data and local target f i (x) Calculate local gradient Then it is sent to the neighboring nodes that can be communicated with. After any node receives the gradient information sent by its neighbor, it can update the parameter x of the local AI model according to the following formula:

[0322]

[0323] Among them, N iis the set of neighbor nodes of node i, |N i | represents the number of elements in the neighbor node set of node i, that is, the number of neighbor nodes of node i, the superscript k represents the kth (k is a positive integer) round of training, α k is the training step size used in the kth round of training. Through information interaction between nodes, the decentralized learning system will eventually learn a unified model.

[0324] However, in the decentralized system described above, due to the lack of unified scheduling by a central node, how different distributed nodes process AI models after receiving them from other nodes is a technical problem that needs to be solved urgently. For example, the AI ​​models used by different distributed nodes may have version gaps during training. For example, some nodes process their local AI models based on newer (or more iterated) AI models, while other nodes process their local AI models based on older (or fewer iterated) AI parameters. This will result in inconsistent AI model versions being obtained by different nodes in the distributed system, affecting the performance of the AI ​​model obtained by the system.

[0325] In order to solve the above problems, the present application provides an AI model processing method and related equipment, which will be further explained in conjunction with the accompanying drawings below.

[0326] Please refer to Figure 4, which is an interactive schematic diagram of the AI ​​model processing method provided in this application. The method includes the following steps.

[0327] S401. The first node determines a first AI model.

[0328] S402. The first node sends first information, where the first information is used to indicate model information of the first AI model and auxiliary information of the first AI model.

[0329] In this embodiment, the first node sends the first information in step S402. Accordingly, the second node receives N pieces of first information sent from N first nodes in step S402, where N is a positive integer. The first information indicates model information of the first AI model and auxiliary information of the first AI model.

[0330] It should be understood that the model information of the first AI model is used to construct the first AI model. Exemplarily, the model information may include at least one of model parameter information and model structure information.

[0331] Optionally, the structural information of the model may include at least one item such as the number of model layers, the number of neurons in each layer of the model, and the connection relationship between layers in the model.

[0332] In one possible implementation, the first AI model determined by the first node in step S401 is a model understandable by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2. Accordingly, the first AI model indicated by the first information sent by the first node in step S402 is a model understandable by the M nodes in the system where the first node is located. This allows other nodes to understand the first AI model and further perform model processing after receiving the first information, even if there are multiple different node types among the M nodes.

[0333] Optionally, the first AI model determined by the first node in step S401 may not be a model that is understandable to some or all of the M nodes in the system where the first node is located. Accordingly, after receiving the model information of the first AI model, some or all of the M nodes may send the model information of the first AI model to other nodes (such as servers, controllers, network devices, central nodes, etc.), so that the other nodes perform model processing based on the model information of the first AI model to obtain and send model information of an understandable model to some or all of the M nodes, so that some or all of the M nodes can perform subsequent model processing based on the model information of the understandable model.

[0334] Optionally, the first AI model determined by the first node in step S401 may not be a model understandable by some or all of the M nodes in the system where the first node resides. Accordingly, after receiving the model information of the first AI model, some or all of the M nodes may discard models that they do not understand.

[0335] It should be understood that the first AI model is a model that can be understood by M nodes in the system where the first node is located. It can be expressed as the first AI model is a model common to the M nodes in the system where the first node is located, the model structure of the first AI model is a model structure that can be understood by the M nodes in the system where the first node is located, the model structure of the first AI model is a model structure common to the M nodes in the system where the first node is located, the first AI model is a public model of the system where the first node is located, or the model structure of the first AI model is a public model structure of the system where the first node is located.

[0336] For example, consider a public AI model, which is understood by all M nodes in the system where the first node resides. In contrast, each of the M nodes in the system where the first node resides may locally process an AI model that is incomprehensible to other nodes. This model is referred to as a local AI model. The following describes, for each of the M nodes, the process of obtaining the public and local AI models.

[0337] 1. The process of acquiring the local AI model.

[0338] Among the M nodes in the system where the first node is located, the local AI model of each node is used for the learning task of the node, and the local AI model structure and parameter quantity of different nodes may be different.

[0339] In one possible implementation, each of the M nodes in the system where the first node is located can select a suitable local AI model based on capability information (i.e., the device capabilities of the device itself) and / or requirement information (i.e., the requirements of the local task).

[0340] Specifically, the capability information (i.e., the device capability of the device itself) mainly includes computing capability and storage capability. Among them, for the computing capability, the computing capability is generally determined by the computing module in the device, and the computing module may include at least one of the central processing units (CPU), graphics processing units (GPU), neural processing units (NPU) or tensor processing units (TPU). For example, in units of floating-point operations per second (flops), it describes the speed at which the device can perform model training and reasoning. For the storage capability, the storage capability is generally determined by the storage module in the device (such as video memory / cache / memory, etc.), in units of bytes (or KB / MB / GB based on bytes), which describes how large a model the device can store.

[0341] Optionally, the demand information (i.e., the requirements of the local task) generally refers to the performance requirements when the node uses a certain model to complete a specific local task, such as positioning accuracy in positioning tasks, classification accuracy in classification tasks, mean square error in information reconstruction tasks, compression rate in compression tasks, etc.

[0342] From the above content, it can be seen that when configuring the local AI model, each of the M nodes in the system where the first node is located can select a model that meets the local task requirements and has the minimum computing power and storage capacity requirements as the local AI model. For example, when multiple models can meet the local task requirements, select the model with the smallest model parameters and the smallest model training / inference calculation amount; or select the model with the best performance as the local AI model among the models whose computing power and storage capacity requirements are less than or equal to the node computing power and storage capacity.

[0343] Optionally, each of the M nodes in the system where the first node is located can also configure a local AI model in a pre-configured manner. For example, the system or standard can provide a set of optional model libraries for a given task for each node to select a local AI model from.

[0344] Furthermore, the configuration of the local AI model can be performed by the management node in the system where the first node is located and then distributed or installed to each node, or it can be performed by each node independently. In the former case, each node can also report its own computing power, storage capacity, and task requirements to the management node.

[0345] 2. The process of obtaining public AI models.

[0346] In the M nodes of the system where the first node resides, a common AI model is used to transfer knowledge between these nodes. Different nodes in the system use the same common AI model with the same structure and parameters. This homogeneous common AI model design simplifies the design of the mechanism for model transmission between nodes (for the implementation of model transmission, see Figure 3b and the related description).

[0347] Similarly, the public AI model can be implemented in a configuration or pre-configured manner for other nodes. For example, it can be pre-given by a management node or standard, and different learning tasks (such as channel state information (CSI) compression, beam management, positioning, etc.) can use different public AI models.

[0348] In addition, the configuration of the public AI model needs to take into account the computing power and / or storage capacity of each node in the system where the first node is located. For example, when the management node determines the public AI model, the management node can collect the computing power and storage capacity information of each node and determine the public AI model based on the computing power and storage capacity of the weakest node. For another example, when the public AI model is pre-given in a pre-configured manner, the standard should also specify the computing power requirements and storage capacity requirements corresponding to the public AI model. Before each node joins the system, it should ensure that its corresponding capacity is not less than the capacity requirements given by the standard.

[0349] Optionally, in the above implementation process, the local AI model configuration and the public AI model configuration are generally performed at the beginning of system construction. When the system status (such as the system environment, the capabilities and requirements of each node) changes, the local AI model and the public AI model can be reconfigured or reset.

[0350] In one possible implementation, the aforementioned local AI model and public AI model can be designed for the same functionality. For example, the local AI model and public AI model can both perform the same functions, such as positioning, classification, reconstruction, compression, and other tasks. In this case, the only differences between the two models are model structure, number of parameters, and performance.

[0351] In another possible implementation, the local AI model and public AI model mentioned above can serve different functions: that is, the functions of the local AI model and the public AI model can be different. For example, if each node generates data based on a GAN architecture, the local AI model can be the generator model in the GAN architecture, while the public AI model can be the discriminator model in the GAN architecture. Specifically, each node trains and updates its local generator and discriminator models based on local data and the discriminator models sent by neighboring nodes, and then sends the updated discriminator model to its neighbors. The discriminator model contains information related to each node's local data, and the interactive discriminator model enables knowledge exchange between nodes. For another example, the local AI model and the public AI model can be two parts of a complete model, such as the local AI model being the encoder portion of an autoencoder model, and the public AI model being the decoder portion of the autoencoder model. Each node can use local data to train a different encoder (local AI model) and the same decoder (public AI model), and then send the decoder (public AI model) to its neighbors. The decoder part also carries the local data information of each node, thus realizing the interaction of knowledge between nodes.

[0352] Exemplarily, as shown in FIG5 a , there is an implementation process of “knowledge sharing” and / or “knowledge absorption” between the above-mentioned local AI model and the public AI model.

[0353] For example, "knowledge sharing" refers to acquiring (or updating) a public AI model based on a local AI model. When the local and public AI models target the same functionality, knowledge sharing can be achieved through existing techniques such as knowledge distillation, pruning, and expansion. When the local and public AI models target different functionality, knowledge sharing can be achieved through joint training of the local and public AI models.

[0354] For example, "knowledge absorption" refers to updating a local AI model based on a public AI model. Similarly, when the local and public AI models target the same functionality, knowledge absorption can be achieved through existing techniques such as knowledge distillation, pruning, and expansion. When the local and public AI models target different functionality, knowledge absorption can be achieved through joint training of the local and public AI models.

[0355] In one possible implementation, the first AI model determined by the first node in step S401 is obtained based on the node type of the first node. Specifically, in the decentralized system shown in Figure 3b above, each node uses the same learning model (the parameters are all x), without considering the different capabilities and requirements of each node, and the model fusion adopts the averaging method shown above, and the functions of each node are the same, and the functional distinction of the nodes is not considered in the above system. In this embodiment, the first AI model indicated by the first information sent by the first node can be a model obtained by the first node based on the node type of the first node. In this case, when the first AI model is obtained at least based on the node type of the first node, the first node can perform different AI model processing processes based on different node types, thereby solving the problem of single node function and improving flexibility.

[0356] Optionally, the first AI model determined by the first node in step S401 is derived based on the node type of the first node and the second AI model. If the first AI model is derived based at least on the second AI model, the first AI model indicated by the first information sent by the first node may be understandable to other nodes, facilitating further model processing by the other nodes after receiving the first AI model.

[0357] Optionally, the node type includes any of the following: a node type that performs local training based on local data (denoted as an L node for ease of description below), a node type that performs fusion processing based on the AI ​​model of other nodes (denoted as an A node for ease of description below), and a node type that performs local training based on the local data and performs fusion processing based on the AI ​​model of other nodes (denoted as an H node for ease of description below). Specifically, the node type used to obtain the first AI model can be implemented by any of the above items to enhance the flexibility of solution implementation.

[0358] Optionally, in the method shown in FIG4 , the method further includes: the first node sending indication information indicating the node type of the first node. Specifically, the first node may further send indication information indicating the node type of the first node, so that other nodes can clearly identify the node type of the first node based on the indication information, and subsequently interact with the first node based on the node type of the first node.

[0359] Optionally, in the method shown in FIG4 , before step S401, the first node determines the node type of the first node based on capability information and / or requirement information; or, the first node receives indication information indicating the node type of the first node. Specifically, the first node may determine the node type of the first node based on its own capability information and / or requirement information, or may determine the node type of the first node based on indications from other nodes, thereby increasing the flexibility of solution implementation.

[0360] For example, the above three node types and their implementation processes will be further described below in conjunction with some implementation examples.

[0361] For L nodes, they can be called local training nodes. Such nodes have the ability and / or demand to share knowledge, but not the ability and / or demand to absorb knowledge. At the same time, there is local data on the node, which can be used for local training.

[0362] For node A, it can be called a model fusion node. This type of node has no ability / need to share knowledge, but has the ability and / or need to absorb knowledge. At the same time, the node may not have local data for local training.

[0363] H nodes can be called hybrid function nodes. Such nodes have both the ability / demand for knowledge sharing and the ability and / or demand for knowledge absorption. At the same time, the nodes have local data that can be used for local training.

[0364] Optionally, similar to the configuration methods of the aforementioned local AI model and public AI model, the type of node can be determined by the node based on its own capabilities, needs and local data conditions, or it can be specified by the management node, or it can be determined by pre-configuration.

[0365] Furthermore, after a node's type is determined, it needs to communicate its node type (i.e., role information) to its neighbors, a process known as role interaction. Furthermore, a node can also communicate its node type along with the type of learning tasks it participates in. For example, after receiving the node type from a neighbor, each node can construct a neighbor information table as shown in Table 1 below.

[0366] Table 1

[0367]

[0368] In the neighbor information table shown in Table 1 above, the node ID is the unique identifier of the neighbor node in the decentralized learning system and can be uniformly assigned when the neighbor node joins the learning system; the neighbor role is determined by the node type sent by the neighbor; the task type indicates the type of learning task in which the neighbor participates, such as CSI compression, beam management, positioning, etc., and the task type can also be determined by the node type sent by the neighbor; the link quality indicator describes the quality of the communication link between a given neighbor and the current node, and can include indicators such as the signal-to-noise ratio (SNR), throughput, latency, and packet loss rate of the communication link.

[0369] In one possible implementation, the roles of individual nodes are not static. During the learning process (or the operation of the learning system), node roles may change due to changes in the communication network topology, node requirements, and environmental conditions. Figure 5b shows some examples of node role transitions.

[0370] For example, the H node cannot obtain new local data, or based on privacy or other considerations, it turns off the knowledge sharing function and transforms into an A node.

[0371] For example, the H node determines that the local data is sufficient to train a local AI model with better performance, turns off the knowledge absorption function, and transforms into an L node.

[0372] For example, if the L node cannot obtain new local data, it turns off the local training function, turns on the knowledge absorption function, and transforms into an A node.

[0373] For example, the L node determines that a local AI model with satisfactory performance cannot be trained based on local data, so it turns on the knowledge absorption function and transforms into an H node.

[0374] For example, node A can obtain local data, enable local training, and transform into node H.

[0375] For example, node A can obtain local data and judge that the local data is sufficient to train a local AI model with better performance. It turns off the knowledge absorption function and transforms into node L.

[0376] Similar to the implementation process of Table 1, after the role of each node changes, the neighboring nodes may be notified of the change so that the neighboring nodes can update the neighbor information tables they maintain.

[0377] In one possible implementation, after a node completes building its neighbor information table, it begins distributed learning by performing relevant operations based on its role. This learning process occurs periodically, with each cycle consisting of two phases: model update and model interaction. Table 2 below lists the specific operations performed by different nodes in each phase.

[0378] Table 2

[0379]

[0380] In the above operations, the local training and updating of the model can be implemented using gradient backpropagation training similar to that used in the fully connected neural network described in Figure 2a above, or can be implemented through other AI training processes, which are not limited here. In addition, the method of obtaining a public AI model based on a local AI model and updating the local AI model based on the public AI model can be referred to the introduction in Figure 5a and the related description above, and will not be repeated here.

[0381] As can be seen from the above implementation process, when the first AI model indicated by the first information sent by the first node is obtained based on a second AI model, the second AI model can be obtained by the first node based on local data; or, the second AI model can be obtained by the first node based on K pieces of information, each of which indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, where K is a positive integer; or, the second AI model can be obtained by the first node based on the local data and the K pieces of information. Specifically, the second AI model used to obtain the first AI model can be implemented using any of the above methods to enhance the flexibility of the solution implementation.

[0382] Optionally, the value of K is 1 less than the value of M, or the value of K is less than M-1.

[0383] In one possible implementation, the auxiliary information of the first AI model includes at least one of the following: type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information of generating the first AI model, geographic location information of generating the first AI model, and distribution information of local data of the first node. Specifically, the auxiliary information of the first AI model indicated by the first information includes at least one of the above items, so that the recipient of the first information can perform AI model processing on the model information of the first AI model based on the at least one information item, thereby improving the performance of the AI ​​model obtained by the recipient of the first information based on the processing of the first AI model.

[0384] Optionally, the model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, which include time domain resources and / or frequency domain resources. Specifically, since the amount of data corresponding to the model information of the first AI model and the amount of data corresponding to the auxiliary information of the first AI model are generally different (for example, the former is generally greater than the latter), the two can be carried on different transmission resources.

[0385] Optionally, the transmission resource is a preconfigured resource.

[0386] For example, as described above, the M nodes in the system where the first node resides send the updated public AI model of the first node to neighboring nodes (for example, the first information sent by the first node to the second node in step S402 carries the model information of the first AI model). Because the structure and parameters of the public AI model are preset and unified, a communication mechanism with low complexity can be designed. The implementation process of the first information sent by the first node in step S402 will be described below with reference to more implementation examples.

[0387] For the first information sent by the first node in step S402, the first AI model indicated by the first information may be a public AI model. The first node may package the first information into a data packet as shown in FIG5c , wherein the model information of the public AI model is carried in the payload, and the auxiliary information of the public AI model is carried in the header. As shown in FIG5c , the header includes at least one item of information, such as the payload type, node ID, version number, timestamp, geographic location, and data distribution.

[0388] Optionally, the payload type is used to indicate the type of public AI model information carried by the payload part, which may be model parameters, model gradients, (intermediate) reasoning results, etc. A bit mapping table can be designed to implicitly indicate this information, for example, 00 represents model parameters, 11 represents model gradients, and 01 or 10 represents (intermediate) reasoning results.

[0389] Optionally, the node ID may include a source node ID and a destination node ID, which is the same as the node ID used in the neighbor information table. The source node ID is used to identify the node that generates the public AI model information contained in the payload, and the destination node ID is used to identify the destination node of the public AI model information contained in the payload.

[0390] Optionally, the version number is used to indicate the version of the public AI model contained in the payload. The design of the version number will be discussed in detail in Example 2.

[0391] Optionally, the timestamp is the time point when the public AI model included in the payload is updated.

[0392] Optionally, the geographic location identifies the geographic location information of the node when the public AI model contained in the payload is updated.

[0393] Optionally, data distribution refers to the distribution of local data on the source node. For data packets generated by nodes without local data, this field can be left blank (or filled with 0). Information such as version number, timestamp, geographic location, and data distribution is used for model fusion and is therefore collectively referred to as model fusion auxiliary information.

[0394] Regarding the first information transmission process, the first node may complete data packet packaging in the above manner and then transmit the data packet to the adjacent node in step S402. For example, the communication link between the first node and the second node is a sidelink, i.e., the data packet is transmitted via the sidelink.

[0395] Optionally, the header and payload of a data packet can be sent separately. For example, the header is sent on a control channel and the payload is sent on a data channel. In this case, the control channel where the header is located also needs to send a transmission resource (time-frequency-space resource) position indicating the data channel used to send the payload corresponding to the header. It should be noted that since the structure and parameter quantity of the public AI model are preset and unified, the amount of transmission resources it uses is also predictable and fixed. Therefore, resources specifically used for the transmission of the public AI model can be pre-allocated, thereby simplifying the design of the communication mechanism (i.e., there is no need to determine the transmission resource allocation and scheduling strategy based on different data packet lengths).

[0396] S403. The second node performs model processing based on the N first information to obtain a target AI model.

[0397] In this embodiment, after the second node receives N first information in step S402, the second node performs model processing based on the N first information in step S403 to obtain a target AI model.

[0398] Optionally, the target AI model is used to complete the AI ​​task of the second node, or the target AI model is a local model of the second node.

[0399] In one possible implementation, the second node performing model processing based on the N first information to obtain the target AI model includes: the second node performing model processing based on the N first information and the node type of the second node to obtain the target AI model. Specifically, when the second AI model is obtained based at least on the node type of the second node, the second node can perform different AI model processing processes based on different node types, thereby resolving the issue of single node functionality and improving flexibility.

[0400] Optionally, the value of K is 1 less than the value of M, or the value of K is less than M-1.

[0401] Exemplarily, in step S403, after the second node receives the data packet containing public AI model information sent by N neighboring points, it can unpack it and obtain the header and payload part. The node fuses the public AI model based on the model fusion auxiliary information (version number, timestamp, geographic location, data distribution, etc.). The public AI model information sent by N neighboring points can be grouped and fused in a multi-level fusion manner. For example, the public AI model information in the data packets with similar timestamps can be grouped together, or the public AI model information in the data packets with similar geographic location information can be grouped together, or the public AI model information in the data packets with different data distributions can be grouped together. Model fusion can be achieved by weighted averaging, distillation and other methods, which will not be described here.

[0402] The following is an illustrative description of the process in which the second node receives N first messages containing the version number information of the AI ​​model. The version number can be designed as AxLyHz, where A, L, and H are fixed fields and x, y, and z are integers. In addition, the version number update rules are as follows: the L node performs a training update on the public AI model based on the local AI model, and the y field in the model version number is increased by 1; the A node obtains an updated public AI model based on the public AI model sent by the neighboring node, and the x field in the model version number is increased by 1 each time the x field is increased by 1, and the y and z fields are set to 0; the H node updates the public AI model based on the local AI model and the public AI model sent by the neighboring node, and the x, y, and z fields in the model version number are increased by 1 respectively, and the x field is increased by 1 each time, and the y and z fields are set to 0.

[0403] As shown in Figure 5d, a circle represents an AI model, and the version number of a certain model is AxLyHz. In the example shown in Figure 5d, the value of x can be an integer such as 1, 2, 3 or 4; the value of y can be an integer such as 0, 1, y, u; the value of z can be an integer such as 0, x, v, w or z. In addition, in the model fusion process based on version numbers shown in Figure 5d, public AI models with similar version numbers can be grouped together, and intra-group fusion is performed first. After completion, grouping and intra-group fusion are performed again until the fusion process is completed. For example, as shown in Figure 5d, public AI models with the same x field are grouped together. Each time an intra-group fusion is completed, the x field is increased by 1, and the y and z fields are reset to 0 until a fused public AI model is obtained.

[0404] It should be noted that the version numbers of the public models finally fused by different nodes in the system where the second node resides are similar, and may even be the same. The purpose of the entire system is to ensure that each node obtains a model for completing local AI tasks. Therefore, the existence of version numbers only allows more similar models to be fused first. It does not necessarily mean that models with higher versions will have better performance.

[0405] Similarly, the second node can also perform fusion (distillation, pruning, expansion) based on auxiliary information of other AI models (such as timestamps, geographic locations, etc.). The specific implementation process can refer to the above example and will not be repeated here.

[0406] In one possible implementation, the process in step S403 whereby the second node performs model processing based on the N first information and the node type of the second node to obtain the target AI model includes: when the node type of the second node is a node type that performs fusion processing based on AI models of other nodes, the second node performs model fusion on the N first AI models based on the N first information to obtain the target AI model. Specifically, when the node type of the second node is a node type that performs fusion processing based on AI models of other nodes, after receiving the N first information and determining the N first AI models, the second node performs model fusion on the N first AI models to obtain the target AI model.

[0407] In one possible implementation, the second node performs model processing based on the N first information and the node type of the second node in step S403 to obtain the target AI model, including the following steps: when the node type of the second node is a node type that performs local training based on the local data and performs fusion processing based on the AI ​​models of other nodes, the second node performs model fusion on the N first AI models and the second AI model based on the N first information to obtain the target AI model, wherein the second AI model is obtained by training based on local data. Specifically, when the node type of the second node is a node type that performs local training based on the local data and performs fusion processing based on the AI ​​models of other nodes, after the second node receives the N first information to determine the N first AI models, the second node needs to obtain the second AI model based on local data training, and perform model fusion on the N first AI models and the second AI model to obtain the target AI model.

[0408] It should be understood that the second AI model used by the second node can refer to the relevant description of the second AI model used by the aforementioned first node and achieve corresponding technical effects, which will not be repeated here.

[0409] In one possible implementation, the method further includes: the second node receiving indication information indicating the node type of the first node. Specifically, the second node may also receive indication information indicating the node type of the first node, so that the second node can determine the node type of the first node based on the indication information, and the second node can subsequently interact with the first node based on the node type of the first node.

[0410] In one possible implementation, the method further includes: the second node sending indication information indicating a node type of the second node. Specifically, the second node may also send indication information indicating the node type of the second node, so that other nodes can determine the node type of the second node based on the indication information and subsequently interact with the second node based on the node type of the second node.

[0411] In one possible implementation, the method further includes: the second node determining the node type of the second node based on capability information and / or requirement information; or, the second node receiving indication information indicating the node type of the second node. Specifically, the second node may determine the node type of the second node based on its own capability information and / or requirement information, or may determine the node type of the second node based on indications from other nodes, thereby increasing the flexibility of solution implementation.

[0412] It should be understood that the relevant implementation process of the node type of the second node (such as determination of the node type, indication of the node type, etc.) can refer to the relevant description of the node type of the aforementioned first node and achieve the corresponding technical effects, which will not be repeated here.

[0413] Based on the technical solution shown in Figure 4, after determining the first AI model, the first node sends first information indicating the model information of the first AI model and auxiliary information of the first AI model. Compared to a method in which different nodes only exchange their respective AI models, because the first information indicates not only the model information of the first AI model but also the auxiliary information of the first AI model, the recipient of the first information can perform AI model processing (e.g., training, fusion, etc.) on the model information of the first AI model based on the auxiliary information of the first AI model, thereby improving the performance of the AI ​​model obtained by the recipient of the first information based on the first AI model.

[0414] Please refer to Figure 6, which is another interactive schematic diagram of the AI ​​model processing method provided in this application. The method includes the following steps.

[0415] S601. The first node obtains a local AI model.

[0416] In this embodiment, the first node obtains a local AI model in step S601, and the local AI model is used to complete the AI ​​task of the first node.

[0417] In one possible implementation, the local AI model obtained by the first node in step S601 is obtained based on local data; or, the local AI model obtained by the first node in step S601 is obtained based on K pieces of information, each of which indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, where K is a positive integer; or, the local AI model obtained by the first node in step S601 is obtained based on the local data and the K pieces of information. Specifically, the local AI model used to obtain the public AI model can be implemented using any of the above methods to enhance the flexibility of the solution implementation.

[0418] It should be understood that the relevant implementation process of the local AI model obtained by the first node in step S601 (such as the configuration of the local AI model, the model processing process of the local AI model, etc.) can refer to the description of the aforementioned Figure 4 and related embodiments, and achieve the corresponding technical effects, which will not be repeated here.

[0419] S602. The first node determines a public AI model based on the local AI model.

[0420] In this embodiment, after the first node obtains the local AI model in step S601, the first node determines a public AI model based on the local AI model in step S402, where the public AI model is a model that can be understood by M nodes in the system where the first node is located, and M is an integer greater than or equal to 2.

[0421] It should be understood that the public AI model determined by the first node in step S602 is a model understandable by M nodes in the system where the first node is located. It can be expressed as the public AI model is a model common to M nodes in the system where the first node is located, the model structure of the public AI model is a model structure understandable by M nodes in the system where the first node is located, and the model structure of the public AI model is a model structure common to M nodes in the system where the first node is located.

[0422] It should be understood that the relevant implementation process of the public AI model (such as the configuration of the public AI model, the model processing process of the public AI model, etc.) can refer to the above-mentioned Figure 4 and the description of the relevant embodiments, and achieve the corresponding technical effects, which will not be repeated here.

[0423] In one possible implementation, the process of the first node determining the public AI model based on the local AI model in step S602 includes: the first node determining the public AI model based on the local AI model and the node type of the first node. Specifically, the first node may determine the public AI model based on the local AI model and the node type of the first node, so that the first node can perform different AI model processing processes based on different node types, thereby resolving the problem of single node functionality and improving flexibility.

[0424] It should be understood that the relevant implementation process of the node type of the first node (such as determination of the node type, indication of the node type, etc.) can refer to the description of the aforementioned Figure 4 and related embodiments, and achieve corresponding technical effects, which will not be repeated here.

[0425] Optionally, the node type includes any of the following: a node type that performs local training based on local data, a node type that performs fusion processing based on AI models of other nodes, or a node type that performs local training based on the local data and performs fusion processing based on AI models of other nodes. Specifically, the node type used to obtain the public AI model can be any of the above to enhance the flexibility of the solution implementation.

[0426] Optionally, in the method shown in FIG6 , the method further includes: the first node sending indication information indicating the node type of the first node. Specifically, the first node may further send indication information indicating the node type of the first node, so that other nodes can clearly identify the node type of the first node based on the indication information, and subsequently interact with the first node based on the node type of the first node.

[0427] Optionally, in the method shown in FIG6 , the method further includes: the first node determining the node type of the first node based on capability information and / or requirement information; or, the first node receiving indication information indicating the node type of the first node. Specifically, the first node may determine the node type of the first node based on its own capability information and / or requirement information, or may determine the node type of the first node based on indications from other nodes, thereby improving the flexibility of solution implementation.

[0428] S603. The first node sends first information, where the first information is used to indicate model information of the public AI model.

[0429] In this embodiment, the first node sends first information indicating model information of a public AI model in step S603. Correspondingly, for the second node, the second node receives N first information sent from N first nodes in step S603, where N is a positive integer.

[0430] It should be understood that the relevant implementation process of the first information (such as the packaging process of the first information, the sending process of the first information, etc.) can refer to the description of the above-mentioned Figure 4 and related embodiments, and achieve the corresponding technical effects, which will not be repeated here.

[0431] S604. The second node updates the local AI model based on the N first information to obtain updated local AI model information.

[0432] In this embodiment, after the second node receives N pieces of first information in step S603, the second node updates the local AI model based on the N pieces of first information in step S604 to obtain updated local AI model information.

[0433] In one possible implementation, the first information sent by the first node in step S603 also indicates the auxiliary information of the public AI model. Specifically, the first information sent by the first node also indicates the auxiliary information of the public AI model. Compared with the method in which different nodes only interact with each other's respective AI models, since the first information indicates the model information of the public AI model, the first information can also indicate the auxiliary information of the public AI model, so that the recipient of the first information can perform AI model processing (such as training, fusion, etc.) on the model information of the public AI model based on the auxiliary information of the first AI model, thereby improving the performance of the AI ​​model obtained by the recipient of the first information based on the public AI model.

[0434] Optionally, the auxiliary information of the public AI model includes at least one of the following: type information of the public AI model, identification information of the first node, identification information of the receiving node of the public AI model, version information of the public AI model, time information of generating the public AI model, geographic location information of generating the public AI model, and distribution information of local data of the public node. Specifically, the auxiliary information of the public AI model indicated by the first information includes at least one of the above items, so that the recipient of the first information can perform AI model processing on the model information of the public AI model based on the at least one information item, thereby improving the performance of the AI ​​model obtained by the recipient of the first information based on the processing of the public AI model.

[0435] Optionally, the model information of the public AI model and the auxiliary information of the public AI model are respectively carried on different transmission resources, which include time domain resources and / or frequency domain resources. Specifically, since the amount of data corresponding to the model information of the public AI model and the amount of data corresponding to the auxiliary information of the public AI model are generally different (for example, the former is generally more than the latter), the two can be carried on different transmission resources.

[0436] It should be understood that the relevant implementation process of the first information (for example, the process of receiving the first information, the process of the second node performing AI model processing based on the first information, etc.) can refer to the description of the aforementioned Figure 4 and related embodiments, and achieve the corresponding technical effects, which will not be repeated here.

[0437] In one possible implementation, in step S604, the process of the second node updating the local AI model based on the N first information to obtain the updated local AI model includes: the second node updating the local AI model based on the N first information and the node type of the second node to obtain the updated local AI model. Specifically, the second node may determine the public AI model based on the N first information and the node type of the second node, so that the second node can perform different AI model processing based on different node types, thereby resolving the problem of single node functionality and improving flexibility.

[0438] Optionally, the node type includes any of the following: a node type that performs fusion processing based on the AI ​​models of other nodes, or a node type that performs local training based on the local data and performs fusion processing based on the AI ​​models of other nodes. Specifically, the node type used to obtain the updated local AI model can be any of the above to improve the flexibility of the solution implementation.

[0439] Optionally, when the node type is a node type that performs fusion processing based on the AI ​​model of other nodes, the local AI model is obtained based on P pieces of information, each of the P pieces of information indicates model information of the AI ​​model of other nodes and auxiliary information of the AI ​​model of other nodes, and P is a positive integer.

[0440] Optionally, when the node type is a node type that performs local training based on the local data and performs fusion processing based on the AI ​​models of other nodes, the local AI model is obtained based on the local data (for example, during the first iterative processing process), or the local AI model is obtained based on the local data and the P information (for example, during other iterative processing processes other than the first iterative processing process), each of the P information indicates the model information of the AI ​​model of the other node and the auxiliary information of the AI ​​model of the other node, and P is a positive integer.

[0441] Specifically, in the case of different node types, the local AI model can be implemented in different ways as described above to enhance the flexibility of the solution implementation.

[0442] Optionally, the method further includes: the second node sending indication information indicating the node type of the second node. Specifically, the second node may also send indication information indicating the node type of the second node, so that other nodes can clearly identify the node type of the second node based on the indication information, and subsequently interact with the second node based on the node type of the second node.

[0443] Optionally, the method further includes: the second node determining the node type of the second node based on the capability information and / or requirement information; or, the second node receiving indication information indicating the node type of the second node. Specifically, the second node may determine the node type of the second node based on its own capability information and / or requirement information, or may determine the node type of the second node based on indications from other nodes, thereby increasing the flexibility of solution implementation.

[0444] It should be understood that the relevant implementation process of the node type of the second node (such as determination of the node type, indication of the node type, etc.) can refer to the description of the aforementioned Figure 4 and related embodiments, and achieve the corresponding technical effects, which will not be repeated here.

[0445] Based on the technical solution shown in Figure 6, a first node obtains a local AI model used to complete its AI task. Based on the local AI model, the first node determines a public AI model. The first node then transmits first information indicating model information of the public AI model. The public AI model is understandable by M nodes in the system in which the first node resides. In other words, the public AI model indicated by the first information transmitted by the first node is understandable by M nodes in the system in which the first node resides. This allows other nodes, even if there are multiple different node types among the M nodes, to understand the first AI model and perform further model processing after receiving the first information.

[0446] Please refer to Figure 7, which is another interactive schematic diagram of the AI ​​model processing method provided in this application. The method includes the following steps.

[0447] S701. The first node determines the node type of the first AI node.

[0448] In this embodiment, the first node determines its node type in step S701 , and the first node performs at least one of step S702 and step S703 after step S701 .

[0449] It should be understood that if the first node executes step S702 and step S703 after step S701, this application does not limit the order in which step S702 and step S703 are executed. For example, step S702 may be executed first and step S703 may be executed later, or step S703 may be executed first and step S702 may be executed later.

[0450] In one possible implementation, the node type determined by the first node in step S701 includes any of the following: a node type that performs local training based on local data, a node type that performs fusion processing based on AI models of other nodes, or a node type that performs local training based on the local data and performs fusion processing based on AI models of other nodes. Specifically, the node type used to obtain the first AI model can be any of the above to enhance the flexibility of solution implementation.

[0451] Optionally, the method further includes: the first node determining the node type of the first node based on the capability information and / or requirement information; or, the first node receiving indication information indicating the node type of the first node. Specifically, the first node may determine the node type of the first node based on its own capability information and / or requirement information, or may determine the node type of the first node based on indications from other nodes, thereby improving the flexibility of solution implementation.

[0452] S702. The first node sends indication information indicating the node type of the first node.

[0453] In this embodiment, after the first node determines the node type of the first node in step S701, the first node sends indication information indicating the node type of the first node in step S702. Accordingly, the second node receives the indication information indicating the node type of the first node in step S702.

[0454] Specifically, the first node may also send indication information indicating the node type of the first node, so that other nodes can clarify the node type of the first node based on the indication information, and subsequently interact with the first node based on the node type of the first node.

[0455] It should be understood that the relevant implementation process of the node type of the first node (such as determination of the node type, indication of the node type, etc.) can refer to the description of the aforementioned Figure 4 and related embodiments, and achieve corresponding technical effects, which will not be repeated here.

[0456] S703. The first node sends first information, where the first information is used to indicate model information of the first AI model.

[0457] In this embodiment, after determining the node type of the first node in step S701, the first node determines a first AI model based on the node type of the first node. Furthermore, in step S703, the first node transmits first information indicating model information of the first AI model. Accordingly, in step S703, the second node receives N pieces of first information transmitted by the N first nodes.

[0458] In one possible implementation, the first AI model indicated by the first information sent by the first node may be a model obtained by the first node based on the node type of the first node. When the first AI model is obtained based at least on the node type of the first node, the first node can execute different AI model processing procedures based on different node types, thereby resolving the issue of single node functionality and improving flexibility.

[0459] Optionally, the first AI model indicated by the first information sent by the first node may be a model derived by the first node based on the first node's node type and the second AI model. If the first AI model is derived based at least on the second AI model, the first AI model indicated by the first information sent by the first node may be a model understandable by other nodes, facilitating further model processing by the other nodes after receiving the first AI model.

[0460] Optionally, the second AI model is derived based on local data; or, the second AI model is derived based on K pieces of information, each of which indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, where K is a positive integer; or, the second AI model is derived based on the local data and the K pieces of information. Specifically, the second AI model used to derive the first AI model can be implemented using any of the above methods to enhance the flexibility of solution implementation.

[0461] Optionally, the first information is also used to indicate auxiliary information of the first AI model. Specifically, the first information sent by the first node also indicates the auxiliary information of the first AI model. Compared with the method in which different nodes only interact with each other's respective AI models, since the first information indicates the model information of the first AI model, the first information can also indicate the auxiliary information of the first AI model, so that the recipient of the first information can perform AI model processing (such as training, fusion, etc.) on the model information of the first AI model based on the auxiliary information of the first AI model, thereby improving the performance of the AI ​​model obtained by the recipient of the first information based on the first AI model.

[0462] Optionally, the auxiliary information of the first AI model includes at least one of the following: type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information of generating the first AI model, geographic location information of generating the first AI model, and distribution information of local data of the first node. Specifically, the auxiliary information of the first AI model indicated by the first information includes at least one of the above items, so that the recipient of the first information can perform AI model processing on the model information of the first AI model based on the at least one information item, thereby improving the performance of the AI ​​model obtained by the recipient of the first information based on the processing of the first AI model.

[0463] Optionally, the model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, which include time domain resources and / or frequency domain resources. Specifically, since the amount of data corresponding to the model information of the public AI model and the amount of data corresponding to the auxiliary information of the public AI model are generally different (for example, the former is generally more than the latter), the two can be carried on different transmission resources.

[0464] Optionally, the first AI model is a model understandable by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2. Specifically, the first AI model indicated by the first information sent by the first node is a model understandable by the M nodes in the system where the first node is located. This allows other nodes to understand the first AI model and further perform model processing after receiving the first information when there are multiple different node types among the M nodes.

[0465] It should be understood that the relevant implementation process of the first information (such as the packaging process of the first information, the sending process of the first information, the receiving process of the first information, the AI ​​model processing process performed by the second node based on the first information, etc.) can refer to the description of the aforementioned Figure 4 and related embodiments, and achieve the corresponding technical effects, which will not be repeated here.

[0466] Based on the technical solution shown in Figure 7, the first node can determine its node type, and then the first node can perform AI model processing based on the node type. This allows the first node to perform different AI model processing processes based on different node types, solving the problem of single node function and improving flexibility.

[0467] Referring to Figure 8, an embodiment of the present application provides a communication device 800. The communication device 800 can implement the functions of the first node (the first node is a terminal device or a network device) in the above method embodiment, and thus can also achieve the beneficial effects of the above method embodiment. In the embodiment of the present application, the communication device 800 can be the first node, or it can be an integrated circuit or component within the first node, such as a chip. The following embodiments are described using the communication device 800 as the first node as an example.

[0468] In one possible implementation, when the device 800 is used to execute the method executed by the first node in any of the aforementioned embodiments, the device 800 includes a processing unit 801 and a transceiver unit 802; the processing unit 801 is used to determine a first AI model; the transceiver unit 802 is used to send first information, where the first information indicates model information of the first AI model and auxiliary information of the first AI model.

[0469] In one possible implementation, the auxiliary information of the first AI model includes at least one of the following:

[0470] Type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information for generating the first AI model, geographic location information for generating the first AI model, and distribution information of local data of the first node.

[0471] In a possible implementation, the model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0472] In one possible implementation, the first AI model is obtained based on the node type of the first node.

[0473] In a possible implementation, the first AI model is obtained based on the second AI model and the node type of the first node.

[0474] In one possible implementation, the second AI model is obtained based on local data; or, the second AI model is obtained based on K pieces of information, each of the K pieces of information indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, where K is a positive integer; or, the second AI model is obtained based on the local data and the K pieces of information.

[0475] In one possible implementation, the node type includes any of the following: a node type that performs local training based on local data, a node type that performs fusion processing based on the AI ​​model of other nodes, and a node type that performs local training based on the local data and performs fusion processing based on the AI ​​model of other nodes.

[0476] In a possible implementation, the transceiver unit 802 is further configured to send indication information indicating the node type of the first node.

[0477] In one possible implementation, the processing unit 801 is further used to determine the node type of the first node based on capability information and / or requirement information; or, the transceiver unit 802 is further used to receive indication information indicating the node type of the first node.

[0478] In one possible implementation, the first AI model is a model that can be understood by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2.

[0479] In one possible implementation, when the device 800 is used to execute the method executed by the second node in any of the aforementioned embodiments, the device 800 includes a processing unit 801 and a transceiver unit 802; the transceiver unit 802 is used to receive N first information, each of the N first information indicates model information of a first AI model and auxiliary information of the first AI model, where N is a positive integer; the processing unit 801 is used to perform model processing based on the N first information to obtain a target AI model.

[0480] Optionally, the target AI model is used to complete the AI ​​task of the second node, or the target AI model is a local model of the second node.

[0481] In one possible implementation, the auxiliary information of the first AI model includes at least one of the following: type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information of generating the first AI model, geographic location information of generating the first AI model, and distribution information of local data of the first node.

[0482] In a possible implementation, the model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0483] In one possible implementation, the processing unit 801 is specifically used to perform model processing based on the N first information and the node type of the second node to obtain the target AI model.

[0484] In one possible implementation, when the node type of the second node is a node type that performs fusion processing based on the AI ​​models of other nodes, the processing unit 801 is specifically used to perform model fusion on the N first AI models based on the N first information to obtain the target AI model.

[0485] In one possible implementation, when the node type of the second node is a node type that performs local training based on the local data and performs fusion processing based on the AI ​​models of other nodes, the processing unit 801 is specifically used to perform model fusion on the N first AI models and the second AI model based on the N first information to obtain the target AI model, wherein the second AI model is obtained by training based on local data.

[0486] In a possible implementation, the transceiver unit 802 is further configured to receive indication information indicating a node type of the first node.

[0487] In a possible implementation, the transceiver unit 802 is further configured to send indication information indicating the node type of the second node.

[0488] In a possible implementation, the processing unit 801 is further used to determine the node type of the second node based on capability information and / or requirement information; or, the transceiver unit 802 is further used to receive indication information indicating the node type of the second node.

[0489] In one possible implementation, the first AI model is a model that can be understood by M nodes in the system where the second node is located, where M is an integer greater than or equal to 2.

[0490] In one possible implementation, when the device 800 is used to execute the method executed by the first node in any of the aforementioned embodiments, the device 800 includes a processing unit 801 and a transceiver unit 802; the processing unit 801 is used to obtain a local AI model, which is used to complete the AI ​​task of the first node; the processing unit 801 is also used to determine a public AI model based on the local AI model, where the public AI model is a model that can be understood by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2; the transceiver unit 802 is used to send first information, which is used to indicate model information of the public AI model.

[0491] In a possible implementation, the processing unit 801 is specifically configured to determine a public AI model based on the local AI model and the node type of the first node.

[0492] In one possible implementation, the node type includes any of the following: a node type that performs local training based on local data, a node type that performs fusion processing based on the AI ​​model of other nodes, and a node type that performs local training based on the local data and performs fusion processing based on the AI ​​model of other nodes.

[0493] In a possible implementation, the transceiver unit 802 is further configured to send indication information indicating the node type of the first node.

[0494] In a possible implementation, the processing unit 802 is further configured to determine the node type of the first node based on capability information and / or requirement information; or, the transceiver unit 802 is further configured to receive indication information indicating the node type of the first node.

[0495] In one possible implementation,

[0496] The local AI model is obtained based on local data; or

[0497] The local AI model is obtained based on K pieces of information, each of the K pieces of information indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, where K is a positive integer; or

[0498] The local AI model is obtained based on the local data and the K information.

[0499] In one possible implementation, the first information also indicates auxiliary information of the public AI model.

[0500] In one possible implementation, the auxiliary information of the public AI model includes at least one of the following:

[0501] Type information of the public AI model, identification information of the first node, identification information of the receiving node of the public AI model, version information of the public AI model, time information of generating the public AI model, geographic location information of generating the public AI model, and distribution information of local data of the public node.

[0502] In a possible implementation, the model information of the public AI model and the auxiliary information of the public AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0503] In one possible implementation, when the device 800 is used to execute the method executed by the second node in any of the aforementioned embodiments, the device 800 includes a processing unit 801 and a transceiver unit 802; the transceiver unit 802 is used to receive N first information, each of the N first information is used to indicate model information of a public AI model, and the public AI model is a model that can be understood by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2; the processing unit 801 is used to update the local AI model based on the N first information to obtain an updated local AI model, and the local AI model is used to complete the AI ​​task of the second node.

[0504] In a possible implementation, the processing unit 801 is specifically configured to update the local AI model based on the N first information and the node type of the second node to obtain an updated local AI model.

[0505] In one possible implementation, the node type includes any of the following: a node type that performs fusion processing based on the AI ​​model of other nodes, and a node type that performs local training based on the local data and performs fusion processing based on the AI ​​model of other nodes.

[0506] In a possible implementation, the transceiver unit 802 is further configured to send indication information indicating the node type of the second node.

[0507] In a possible implementation, the processing unit 801 is further used to determine the node type of the second node based on capability information and / or requirement information; or, the transceiver unit 802 is further used to receive indication information indicating the node type of the second node.

[0508] In one possible implementation, when the node type is a node type that performs fusion processing based on the AI ​​model of other nodes, the local AI model is obtained based on P pieces of information, each of the P pieces of information indicates model information of the AI ​​model of other nodes and auxiliary information of the AI ​​model of other nodes, and P is a positive integer.

[0509] In one possible implementation, when the node type is a node type that performs local training based on the local data and performs fusion processing based on the AI ​​models of other nodes, the local AI model is obtained based on the local data, or the local AI model is obtained based on the local data and the P information, each of the P information indicates model information of the AI ​​model of other nodes and auxiliary information of the AI ​​model of other nodes, and P is a positive integer.

[0510] In one possible implementation, the first information is also used to indicate auxiliary information of the public AI model.

[0511] In one possible implementation, the auxiliary information of the public AI model includes at least one of the following: type information of the public AI model, identification information of the first node, identification information of the receiving node of the public AI model, version information of the public AI model, time information of generating the public AI model, geographic location information of generating the public AI model, and distribution information of local data of the first node.

[0512] In a possible implementation, the model information of the public AI model and the auxiliary information of the public AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0513] In one possible implementation, when the device 800 is used to execute the method executed by the first node in any of the aforementioned embodiments, the device 800 includes a processing unit 801 and a transceiver unit 802; the processing unit 801 is used by the first node to determine the node type of the first node.

[0514] In a possible implementation, the apparatus further includes a transceiver unit, and the transceiver unit 802 is configured to send indication information indicating the node type of the first node.

[0515] In one possible implementation, the device further includes a transceiver unit 802, which is used to send first information, where the first information is used to indicate model information of the first AI model, wherein the first AI model is obtained based on the node type of the first node.

[0516] Optionally, the first AI model is obtained based on the second AI model and the node type of the first node.

[0517] In one possible implementation,

[0518] The second AI model is obtained based on local data; or

[0519] The second AI model is obtained based on K pieces of information, each of the K pieces of information indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, where K is a positive integer; or

[0520] The second AI model is obtained based on the local data and the K information.

[0521] In one possible implementation, the node type includes any of the following: a node type that performs local training based on local data, a node type that performs fusion processing based on the AI ​​model of other nodes, and a node type that performs local training based on the local data and performs fusion processing based on the AI ​​model of other nodes.

[0522] In one possible implementation, the first information is also used to indicate auxiliary information of the first AI model, wherein the auxiliary information of the first AI model includes at least one of the following: type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information of generating the first AI model, geographic location information of generating the first AI model, and distribution information of local data of the first node.

[0523] In a possible implementation, the model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0524] In one possible implementation, the first AI model is a model that can be understood by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2.

[0525] In one possible implementation, the processing unit 801 is further used to determine the node type of the first node based on capability information and / or requirement information; or, the transceiver unit 802 is further used to receive indication information indicating the node type of the first node.

[0526] In one possible implementation, when the device 800 is used to execute the method executed by the second node in any of the aforementioned embodiments, the device 800 includes a transceiver unit 802; the transceiver unit 802 is used for the second node to receive indication information indicating the node type of the first node; and / or, the transceiver unit 802 is used to receive first information, and the first information is used to indicate model information of the first AI model, wherein the first AI model is obtained based on the node type of the first node.

[0527] Optionally, the first AI model is obtained based on the second AI model and the node type of the first node.

[0528] In one possible implementation,

[0529] The second AI model is obtained based on local data; or

[0530] The second AI model is obtained based on K pieces of information, each of the K pieces of information indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, where K is a positive integer; or

[0531] The second AI model is obtained based on the local data and the K information.

[0532] In one possible implementation, the node type includes any of the following: a node type that performs local training based on local data, a node type that performs fusion processing based on the AI ​​model of other nodes, and a node type that performs local training based on the local data and performs fusion processing based on the AI ​​model of other nodes.

[0533] In one possible implementation, the first information is also used to indicate auxiliary information of the first AI model, wherein the auxiliary information of the first AI model includes at least one of the following: type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information of generating the first AI model, geographic location information of generating the first AI model, and distribution information of local data of the first node.

[0534] In a possible implementation, the model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

[0535] In one possible implementation, the first AI model is a model that can be understood by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2.

[0536] In one possible implementation, the device further includes a processing unit 801, which is used to determine the node type of the second node based on capability information and / or requirement information; or, the transceiver unit is further used to receive indication information indicating the node type of the second node.

[0537] It should be noted that, for details on the information execution process of the units of the above-mentioned communication device 800, please refer to the description in the method embodiment shown above in this application, and no further details will be given here.

[0538] Please refer to Fig. 9, which is another schematic structural diagram of a communication device 900 provided in this application. The communication device 900 at least includes an input and output interface 902. The communication device 900 may be a chip or an integrated circuit.

[0539] Optionally, the communication device further includes a logic circuit 901 .

[0540] The transceiver unit 802 shown in FIG8 may be a communication interface, which may be the input / output interface 902 in FIG9 , which may include an input interface and an output interface. Alternatively, the communication interface may be a transceiver circuit, which may include an input interface circuit and an output interface circuit.

[0541] Optionally, the logic circuit 901 is used to determine a first AI model; the input / output interface 902 is used to send first information, where the first information indicates model information of the first AI model and auxiliary information of the first AI model.

[0542] Optionally, the input-output interface 902 is used to receive N first information, each of the N first information indicating model information of the first AI model and auxiliary information of the first AI model; the logic circuit 901 is used to perform model processing based on the N first information to obtain the target AI model.

[0543] Optionally, the logic circuit 901 is used to obtain a local AI model, which is used to complete the AI ​​task of the first node; the logic circuit 901 is used to determine a public AI model based on the local AI model, where the public AI model is a model that can be understood by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2; the input and output interface 902 is used to send first information, which is used to indicate model information of the public AI model.

[0544] Optionally, the input-output interface 902 is used to receive N first information, each of the N first information is used to indicate model information of a public AI model, where the public AI model is a model understandable by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2; the logic circuit 901 is used to update the local AI model based on the model information of the public AI model to obtain an updated local AI model, which is used to complete the AI ​​task of the second node.

[0545] Optionally, the logic circuit 901 is configured to be used by the first node to determine the node type of the first node.

[0546] Optionally, the input / output interface 902 is used for the second node to receive indication information indicating the node type of the first node;

[0547] Optionally, the input / output interface 902 is used to receive first information, where the first information is used to indicate model information of the first AI model, wherein the first AI model is obtained based on the node type of the first node.

[0548] The logic circuit 901 and the input / output interface 902 may also execute other steps executed by the network device in any embodiment and achieve corresponding beneficial effects, which will not be described in detail here.

[0549] In a possible implementation, the processing unit 801 shown in FIG. 8 may be the logic circuit 901 in FIG. 9 .

[0550] Optionally, the logic circuit 901 may be a processing device, and the functions of the processing device may be partially or entirely implemented by software.

[0551] Optionally, the processing device may include a memory and a processor, wherein the memory is used to store a computer program, and the processor reads and executes the computer program stored in the memory to perform corresponding processing and / or steps in any one of the method embodiments.

[0552] Alternatively, the processing device may include only a processor. A memory for storing the computer program is located outside the processing device, and the processor is connected to the memory via circuits / wires to read and execute the computer program stored in the memory. The memory and processor may be integrated or physically separate.

[0553] Optionally, the processing device may be one or more chips, or one or more integrated circuits. For example, the processing device may be one or more field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), system-on-chips (SoCs), central processor units (CPUs), network processors (NPs), digital signal processors (DSPs), microcontroller units (MCUs), programmable logic devices (PLDs), or other integrated chips, or any combination of the above chips or processors.

[0554] Please refer to Figure 10, which shows the communication device 1000 involved in the above-mentioned embodiments provided in an embodiment of the present application. The communication device 1000 can specifically be a communication device serving as a terminal device in the above-mentioned embodiments. The example shown in Figure 10 is that the terminal device is implemented through the terminal device (or a component in the terminal device).

[0555] Herein, a possible logical structure diagram of the communication device 1000 is shown. The communication device 1000 may include but is not limited to at least one processor 1001 and a communication port 1002 .

[0556] Further optionally, the device may also include at least one of a memory 1003 and a bus 1004. In an embodiment of the present application, the at least one processor 1001 is used to control and process the actions of the communication device 1000.

[0557] In addition, the processor 1001 can be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field programmable gate array or other programmable logic device, a transistor logic device, a hardware component, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. The processor can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a digital signal processor and a microprocessor, and so on. Those skilled in the art will clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0558] It should be noted that the communication device 1000 shown in Figure 10 can be specifically used to implement the steps implemented by the terminal device in the aforementioned method embodiment and achieve the corresponding technical effects of the terminal device. The specific implementation methods of the communication device shown in Figure 10 can refer to the description in the aforementioned method embodiment and will not be repeated here.

[0559] Please refer to Figure 11, which is a structural diagram of the communication device 1100 involved in the above-mentioned embodiments provided in an embodiment of the present application. The communication device 1100 can specifically be a communication device as a network device in the above-mentioned embodiments. The example shown in Figure 11 is that the network device is implemented through the network device (or a component in the network device), wherein the structure of the communication device can refer to the structure shown in Figure 11.

[0560] The communication device 1100 includes at least one processor 1111 and at least one network interface 1114. Further optionally, the communication device also includes at least one memory 1112, at least one transceiver 1113 and one or more antennas 1115. The processor 1111, the memory 1112, the transceiver 1113 and the network interface 1114 are connected, for example, via a bus. In an embodiment of the present application, the connection may include various interfaces, transmission lines or buses, etc., which are not limited in this embodiment. The antenna 1115 is connected to the transceiver 1113. The network interface 1114 is used to enable the communication device to communicate with other communication devices through a communication link. For example, the network interface 1114 may include a network interface between the communication device and the core network device, such as an S1 interface, and the network interface may include a network interface between the communication device and other communication devices (such as other network devices or core network devices), such as an X2 or Xn interface.

[0561] Processor 1111 is primarily used to process communication protocols and communication data, control the entire communication device, execute software programs, and process software program data, for example, to support the communication device in performing the actions described in the embodiments. The communication device may include a baseband processor and a central processing unit. The baseband processor is primarily used to process communication protocols and communication data, while the central processing unit is primarily used to control the entire terminal device, execute software programs, and process software program data. Processor 1111 in Figure 11 may integrate the functions of both a baseband processor and a central processing unit. Those skilled in the art will appreciate that the baseband processor and the central processing unit may also be independent processors interconnected via a bus or other technology. Those skilled in the art will appreciate that a terminal device may include multiple baseband processors to accommodate different network standards, multiple central processing units to enhance its processing capabilities, and various components of the terminal device may be connected via various buses. The baseband processor may also be referred to as a baseband processing circuit or a baseband processing chip. The central processing unit may also be referred to as a central processing circuit or a central processing chip. The functionality for processing communication protocols and communication data may be built into the processor or stored in memory as a software program, which is executed by the processor to implement the baseband processing functionality.

[0562] The memory is primarily used to store software programs and data. Memory 1112 can exist independently and be connected to processor 1111. Alternatively, memory 1112 can be integrated with processor 1111, for example, within a single chip. Memory 1112 can store program code for executing the technical solutions of the embodiments of the present application, and execution is controlled by processor 1111. The various computer program codes executed can also be considered drivers for processor 1111.

[0563] Figure 11 shows only one memory and one processor. In an actual terminal device, there may be multiple processors and multiple memories. The memory may also be referred to as a storage medium or a storage device. The memory may be a storage element on the same chip as the processor, i.e., an on-chip storage element, or an independent storage element, which is not limited in the present embodiment.

[0564] The transceiver 1113 can be used to support the reception or transmission of radio frequency signals between the communication device and the terminal. The transceiver 1113 can be connected to the antenna 1115. The transceiver 1113 includes a transmitter Tx and a receiver Rx. Specifically, one or more antennas 1115 can receive radio frequency signals. The receiver Rx of the transceiver 1113 is used to receive the radio frequency signal from the antenna, convert the radio frequency signal into a digital baseband signal or a digital intermediate frequency signal, and provide the digital baseband signal or digital intermediate frequency signal to the processor 1111 so that the processor 1111 can further process the digital baseband signal or digital intermediate frequency signal, such as demodulation and decoding. In addition, the transmitter Tx in the transceiver 1113 is also used to receive a modulated digital baseband signal or digital intermediate frequency signal from the processor 1111, convert the modulated digital baseband signal or digital intermediate frequency signal into a radio frequency signal, and transmit the radio frequency signal through one or more antennas 1115. Specifically, the receiver Rx can selectively perform one or more stages of down-mixing and analog-to-digital conversion on the RF signal to obtain a digital baseband signal or a digital intermediate frequency signal. The order of the down-mixing and analog-to-digital conversion processes is adjustable. The transmitter Tx can selectively perform one or more stages of up-mixing and digital-to-analog conversion on the modulated digital baseband signal or digital intermediate frequency signal to obtain a RF signal. The order of the up-mixing and digital-to-analog conversion processes is adjustable. The digital baseband signal and the digital intermediate frequency signal may be collectively referred to as digital signals.

[0565] The transceiver 1113 may also be referred to as a transceiver unit, a transceiver, a transceiver device, etc. Optionally, a device in the transceiver unit that implements a receiving function may be referred to as a receiving unit, and a device in the transceiver unit that implements a transmitting function may be referred to as a transmitting unit. That is, the transceiver unit includes a receiving unit and a transmitting unit. The receiving unit may also be referred to as a receiver, an input port, a receiving circuit, etc., and the transmitting unit may be referred to as a transmitter, a transmitter, or a transmitting circuit, etc.

[0566] It should be noted that the communication device 1100 shown in Figure 11 can be specifically used to implement the steps implemented by the network device in the aforementioned method embodiment, and to achieve the corresponding technical effects of the network device. The specific implementation methods of the communication device 1100 shown in Figure 11 can refer to the description in the aforementioned method embodiment, and will not be repeated here one by one.

[0567] An embodiment of the present application further provides a computer-readable storage medium storing one or more computer-executable instructions. When the computer-executable instructions are executed by a processor, the processor executes the method described in the possible implementation manner of the terminal device in the aforementioned embodiment.

[0568] An embodiment of the present application further provides a computer-readable storage medium storing one or more computer-executable instructions. When the computer-executable instructions are executed by a processor, the processor executes the method described in the possible implementation of the network device in the aforementioned embodiment.

[0569] An embodiment of the present application also provides a computer program product (or computer program) that stores one or more computers. When the computer program product is executed by the processor, the processor executes the method of the possible implementation method of the above-mentioned terminal device.

[0570] An embodiment of the present application also provides a computer program product storing one or more computers. When the computer program product is executed by the processor, the processor executes the method of the possible implementation of the above-mentioned network device.

[0571] An embodiment of the present application also provides a chip system, which includes at least one processor for supporting a communication device to implement the functions involved in the possible implementation methods of the above-mentioned communication device. Optionally, the chip system also includes an interface circuit, which provides program instructions and / or data to the at least one processor. In one possible design, the chip system may also include a memory, which is used to store the necessary program instructions and data for the communication device. The chip system can be composed of chips, or it can include chips and other discrete devices, wherein the communication device can specifically be the terminal device in the aforementioned method embodiment.

[0572] An embodiment of the present application also provides a chip system, which includes at least one processor for supporting a communication device to implement the functions involved in the possible implementation methods of the above-mentioned communication device. Optionally, the chip system also includes an interface circuit, which provides program instructions and / or data to the at least one processor. In one possible design, the chip system may also include a memory, which is used to store the necessary program instructions and data for the communication device. The chip system can be composed of a chip, or it can include a chip and other discrete devices, wherein the communication device can specifically be the network device in the aforementioned method embodiment.

[0573] An embodiment of the present application also provides a communication system, wherein the network system architecture includes a first node and a second node in any of the above embodiments, wherein the first node may be a terminal device or a network device, and the second node may also be a terminal device or a network device.

[0574] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0575] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0576] In addition, the functional units in the various embodiments of the present application can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit. If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

Claims

1. An artificial intelligence (AI) model processing method, characterized in that: include: The first node determines a first AI model; The first node sends first information, where the first information indicates model information of the first AI model and auxiliary information of the first AI model.

2. The method according to claim 1, characterized in that The auxiliary information of the first AI model includes at least one of the following: Type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information for generating the first AI model, geographic location information for generating the first AI model, and distribution information of local data of the first node.

3. The method according to claim 1 or 2, characterized in that The model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

4. The method according to any one of claims 1 to 3, characterized in that The first AI model is obtained based on the node type of the first node.

5. The method according to any one of claims 1 to 4, characterized in that The first AI model is obtained based on the second AI model and the node type of the first node; The second AI model is obtained based on local data; or The second AI model is obtained based on K pieces of information, each of the K pieces of information indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, and K is a positive integer; or The second AI model is obtained based on the local data and the K pieces of information.

6. The method according to claim 4 or 5, characterized in that The node type includes any of the following: Node types that perform local training based on local data, node types that perform fusion processing based on AI models of other nodes, and node types that perform local training based on the local data and perform fusion processing based on AI models of other nodes.

7. The method according to any one of claims 4 to 6, characterized in that The method further comprises: The first node sends indication information indicating a node type of the first node.

8. The method according to any one of claims 4 to 7, characterized in that The method further comprises: The first node determines the node type of the first node based on capability information and / or requirement information; or, The first node receives indication information indicating a node type of the first node.

9. The method according to any one of claims 1 to 8, characterized in that The first AI model is a model that can be understood by M nodes in the system where the first node is located, where M is an integer greater than or equal to 2.

10. An artificial intelligence (AI) model processing method, characterized in that: include: The second node receives N first information from N first nodes, where each piece of the N first information indicates model information of a first AI model and auxiliary information of the first AI model, and N is a positive integer; The second node performs model processing based on the N first information to obtain a target AI model.

11. The method according to claim 10, characterized in that The auxiliary information of the first AI model includes at least one of the following: type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information for generating the first AI model, geographic location information for generating the first AI model, and distribution information of local data of the first node.

12. The method according to claim 10 or 11, characterized in that The model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

13. The method according to any one of claims 10 to 12, characterized in that The second node performs model processing based on the N first information to obtain a target AI model, including: The second node performs model processing based on the N first information and the node type of the second node to obtain the target AI model.

14. The method according to claim 13, characterized in that The second node performs model processing based on the N first information and the node type of the second node to obtain the target AI model, including: When the node type of the second node is a node type that performs fusion processing based on the AI ​​models of other nodes, the second node performs model fusion on the N first AI models based on the N first information to obtain the target AI model.

15. The method according to claim 13 or 14, characterized in that The second node performs model processing based on the N first information and the node type of the second node to obtain the target AI model, including: When the node type of the second node is a node type that performs local training based on the local data and performs fusion processing based on the AI ​​models of other nodes, the second node performs model fusion on the N first AI models and the second AI model based on the N first information to obtain the target AI model, wherein the second AI model is trained based on local data.

16. The method according to any one of claims 13 to 15, characterized in that The method further comprises: The second node receives indication information indicating the node type of the first node.

17. The method according to any one of claims 13 to 16, characterized in that The method further comprises: The second node sends indication information used to indicate the node type of the second node.

18. The method according to any one of claims 13 to 17, characterized in that The method further comprises: The second node determines the node type of the second node based on capability information and / or requirement information; or, The second node receives indication information indicating a node type of the second node.

19. The method according to any one of claims 10 to 18, characterized in that The first AI model is a model that can be understood by M nodes in the system where the second node is located, where M is an integer greater than or equal to 2.

20. An artificial intelligence (AI) model processing device, characterized in that: including a processing unit and a transceiver unit; The processing unit is used to determine a first AI model; The transceiver unit is used to send first information, where the first information indicates model information of the first AI model and auxiliary information of the first AI model.

21. The device according to claim 20, characterized in that The auxiliary information of the first AI model includes at least one of the following: type information of the first AI model, identification information of the node corresponding to the device, identification information of the receiving node of the first AI model, version information of the first AI model, time information for generating the first AI model, geographic location information for generating the first AI model, and distribution information of local data of the node corresponding to the device.

22. The device according to claim 20 or 21, characterized in that The model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

23. The device according to any one of claims 20 to 22, characterized in that The first AI model is obtained based on the node type corresponding to the device.

24. The device according to claim 23, characterized in that The first AI model is obtained based on the second AI model and the node type; The second AI model is obtained based on local data; or The second AI model is obtained based on K pieces of information, each of the K pieces of information indicates model information of the AI ​​model of another node and auxiliary information of the AI ​​model of the other node, and K is a positive integer; or The second AI model is obtained based on the local data and the K pieces of information.

25. The device according to claim 23 or 24, characterized in that The node type includes any of the following: Node types that perform local training based on local data, node types that perform fusion processing based on AI models of other nodes, and node types that perform local training based on the local data and perform fusion processing based on AI models of other nodes.

26. The device according to any one of claims 23 to 25, characterized in that The transceiver unit is further configured to send indication information indicating a node type corresponding to the device.

27. The device according to any one of claims 23 to 26, characterized in that The processing unit is further configured to determine a node type of a node corresponding to the device based on capability information and / or requirement information; or, The transceiver unit is further configured to receive indication information indicating a node type of a node corresponding to the device.

28. The device according to any one of claims 20 to 27, characterized in that The first AI model is a model that can be understood by M nodes in the system where the node corresponding to the device is located, where M is an integer greater than or equal to 2.

29. An artificial intelligence (AI) model processing device, characterized in that: including a processing unit and a transceiver unit; The transceiver unit is configured to receive N first information from N first nodes, where each piece of the N first information indicates model information of a first AI model and auxiliary information of the first AI model, and N is a positive integer; The processing unit is used to perform model processing based on the N first information to obtain a target AI model.

30. The device according to claim 29, characterized in that The auxiliary information of the first AI model includes at least one of the following: type information of the first AI model, identification information of the first node, identification information of the receiving node of the first AI model, version information of the first AI model, time information for generating the first AI model, geographic location information for generating the first AI model, and distribution information of local data of the first node.

31. The device according to claim 29 or 30, characterized in that The model information of the first AI model and the auxiliary information of the first AI model are respectively carried on different transmission resources, and the transmission resources include time domain resources and / or frequency domain resources.

32. The device according to any one of claims 29 to 31, characterized in that The processing unit is specifically used to perform model processing based on the N first information and the node type corresponding to the processing device to obtain the target AI model.

33. The device according to claim 32, characterized in that When the node type is a node type that performs fusion processing based on AI models of other nodes, the processing unit is specifically configured to perform model fusion on the N first AI models based on the N first information to obtain the target AI model.

34. The device according to claim 32 or 33, characterized in that When the node type is a node type that performs local training based on the local data and performs fusion processing based on the AI ​​models of other nodes, the processing unit is specifically used to perform model fusion on the N first AI models and the second AI model based on the N first information to obtain the target AI model, wherein the second AI model is obtained by training based on local data.

35. The device according to any one of claims 32 to 34, characterized in that The transceiver unit is further configured to receive indication information indicating a node type of the first node.

36. The device according to any one of claims 32 to 35, characterized in that The transceiver unit is further configured to send indication information for indicating a node type corresponding to the device.

37. The device according to any one of claims 32 to 36, characterized in that The processing unit is further configured to determine a node type corresponding to the device based on capability information and / or requirement information; or, The transceiver unit is further configured to receive indication information indicating a node type corresponding to the device.

38. The device according to any one of claims 29 to 37, characterized in that The first AI model is a model understandable by M nodes in the system where the device is located, where M is an integer greater than or equal to 2.

39. A communication device, characterized in that: including at least one logic circuit and an input and output interface; The at least one logic circuit is used to determine a first AI model; The input / output interface is configured to output first information, where the first information indicates model information of the first AI model and auxiliary information of the first AI model; The logic circuit is further configured to execute the method according to any one of claims 1 to 9.

40. A communication device, characterized in that: including at least one logic circuit and an input and output interface; The input / output interface is configured to input N first information, where each piece of the N first information indicates model information of a first AI model and auxiliary information of the first AI model, and N is a positive integer; The at least one logic circuit is configured to perform model processing based on the N first information to obtain a target AI model; The logic circuit is further configured to execute the method according to any one of claims 10 to 19.

41. A communication system, characterized in that The communication system comprises the apparatus according to any one of claims 20 to 28, and the apparatus according to any one of claims 29 to 38; or, The communication system includes the apparatus of claim 39 and the apparatus of claim 40.

42. A computer-readable storage medium, characterized in that The medium stores instructions, and when the instructions are executed by a computer, the method according to any one of claims 1 to 19 is implemented.

43. A computer program product, characterized in that The method comprises instructions which, when executed on a computer, cause the computer to perform the method according to any one of claims 1 to 19.

Images