Network management method and related product

Abstract

Disclosed in the present application are a network management method and a related product. The method comprises: an orchestration agent acquiring an intention request, obtaining a first task on the basis of the intention request and a network model, and sending information of the first task to an assembly agent; the assembly agent obtaining a first set of functional modules on the basis of a first sub-task in the first task and a first tool library; the assembly agent obtaining information of a first network topology on the basis of at least one functional module in the first set of functional modules and a second sub-task in the first task, and sending same to an execution agent; the execution agent obtaining a core network on the basis of the at least one functional module in the first network topology; and a connection agent establishing a communication connection between the core network and a communication device. By using the method in embodiments of the present application, an AI agent can be deployed in a network architecture and various functional services of the AI agent are implemented. The deployment means has low levels of difficulty and high levels of system flexibility and fault tolerance.

Description

Network management method and related products

[0001] This application claims priority to the Chinese Patent Application No. 202410875109.3, filed on July 01, 2024, and entitled "Network management method and related products", the content of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0002] The present application relates to the field of communication technology, in particular to a network management method and related products. BACKGROUND

[0003] With the steady development and rapid popularization of artificial intelligence (AI) technology, especially the development of machine learning (ML), it is expected that intelligence will appear in every part of the communication system to support the construction of smart cities and communities. Therefore, the idea of integrating communication networks with AI technology has been proposed, which can be referred to as network AI.

[0004] Network AI is to provide a complete AI environment and AI services in the network through a unified architecture design. This will bring great changes to the traditional network architecture, evolving from the traditional connection-oriented service-centric architecture to the task-centric architecture for AI new services. For example, by introducing a new dimension of "task" in the future communication network architecture, the quality of service (QoS) of network AI services can be better controlled and guaranteed at the network execution level. This also makes it possible to have a measurable dimension at the network operation level, which can support the new business model of the next-generation network architecture, such as the new business model of the future communication network.

[0005] The core of AI technology is AI models, and the integration of communication networks with AI technology is more about the application of AI models in network architecture. The application of AI models in network architecture is a complex system engineering that requires a series of supporting functions to truly realize its potential. This system engineering presents an AI agent to the outside. Therefore, it is necessary to redesign the network architecture with the idea of AI agent, so as to achieve the goal of regarding the network as an AI-enabled infrastructure that provides intelligent services for various applications.

[0006] How to design network architecture with the idea of AI agent and enable AI in communication networks is a problem that needs to be solved urgently. SUMMARY

[0007] Embodiments of the present application provide a network management method and related products, which provide a way of deploying function entities in a network architecture to implement an AI agent, so that the AI agent can provide AI services for applications in combination with the network architecture. Meanwhile, the deployment method has the advantages of low implementation difficulty, high system flexibility and fault tolerance.

[0008] In a first aspect, the present application provides a network management method. The method is applied to a multi-agent system, and the multi-agent system includes an orchestration agent, an assembly agent, an execution agent, and a connection agent. The orchestration agent obtains an intent request, obtains a first task according to the intent request and a network model, and sends information of the first task to the assembly agent. The first task includes at least one first subtask and at least one second subtask, the first subtask corresponds to a core network, and the second subtask corresponds to a communication device. The assembly agent obtains a first function module set according to the at least one first subtask and a first tool library, and the first function module set includes at least one function module corresponding to the at least one first subtask. The assembly agent obtains a first network topology according to the at least one function module in the first function module set and the at least one second subtask, and sends information of the first network topology to the execution agent. The execution agent receives the information of the first network topology from the assembly agent, obtains a core network corresponding to the first network topology according to the at least one function module in the first network topology, and sends connection-related information to the connection agent. The connection-related information includes information of the core network and task topology information related to the at least one second subtask in the first network topology. The connection agent receives the connection-related information, configures the at least one second subtask on the communication device according to the task topology information, and establishes a communication connection between the core network and the communication device to build an application network. The communication device includes at least one of an access network device or a terminal device.

[0009] Embodiments of the present application construct a multi-agent system. The orchestration agent, the assembly agent, the execution agent, and the connection agent in the multi-agent system are respectively used to implement task orchestration, matching and assembly of tools, deployment and connection of a core network and a communication device, and the like, so as to complete customized application network services through interaction between the multi-agents. The deployment of the system can implement the function services of the AI agent, and meanwhile, the deployment method of the multi-agent reduces the implementation difficulty of the agent and improves the system flexibility and fault tolerance.

[0010] In a possible implementation, before the orchestration agent obtains the first task, the method further includes: the orchestration agent obtaining a second task according to the intention request and the network model, and sending information of the second task to the assembly agent, the second task including at least one first subtask different from the first task; the assembly agent receiving the information of the second task from the orchestration agent, determining that obtaining the second set of function modules according to the at least one first subtask in the second task and the first tool library fails, and sending a first response to the orchestration agent, the first response indicating that obtaining the second set of function modules fails; the orchestration agent receiving the first response, and obtaining the first task according to the intention request and the network model.

[0011] In the embodiments of the present application, when it is determined that the subtask fails to match the function module, a cyclic optimization process between the assembly agent and the orchestration agent is used to implement complete matching of the subtask to the function module, thereby ensuring normal implementation of the user intention.

[0012] In a possible implementation, before the assembly agent obtains the first set of function modules, the method further includes: the assembly agent obtaining a third set of function modules according to the at least one first subtask in the first task and the first tool library, the third set of function modules including at least one function module corresponding to the at least one first subtask; the assembly agent obtaining a second network topology according to the at least one function module in the third set of function modules and the at least one second subtask, and sending information of the second network topology to the execution agent; the execution agent receiving the information of the second network topology from the assembly agent, determining that obtaining the core network corresponding to the second network topology fails according to the at least one function module in the second network topology, and sending a second response to the assembly agent, the second response indicating that obtaining the core network corresponding to the second network topology fails; the assembly agent receiving the second response from the execution agent, and obtaining the first set of function modules according to the at least one first subtask and the first tool library, including: the assembly agent obtaining the first set of function modules according to the second response, the at least one first subtask and the first tool library, wherein the third set of function modules and the first set of function modules include at least one different function module.

[0013] In the embodiments of the present application, when deployment of the function module fails, a cyclic optimization process between the assembly agent and the execution agent is used to implement complete deployment of the function module, thereby ensuring complete implementation of the user intention.

[0014] In one feasible implementation, the executing agent obtains the core network corresponding to the first network topology based on at least one functional module in the first network topology, including: the executing agent deploys at least one functional module in the first network topology to at least one core network element to obtain a core network composed of at least one core network element.

[0015] In one feasible implementation, the first network topology further includes the resource requirements of each functional module in at least one functional module, and deploying at least one functional module in the first network topology to at least one core network element includes: deploying at least one functional module in the first network topology to at least one core network element according to the resource requirements of each functional module.

[0016] In one feasible implementation, the execution agent determines that the core network failure corresponding to the second network topology is obtained based on at least one functional module in the second network topology. This includes: the execution agent determines that the second network topology includes a functional module that has failed to be deployed. The functional module deployment failure includes at least one of the following: the core network does not meet the resource requirements of the functional module; or the resources of the core network do not meet the QoS requirements of the functional module.

[0017] In one feasible implementation, the method further includes: an agent acquiring network status information of the application network; when the network status information indicates that the application network is performing abnormally, the agent adjusts the application network, including at least one of the following: adjusting the quantity and position of at least one functional module; adjusting the quantity and position of at least one second subtask.

[0018] In this application embodiment, a method is disclosed to discover and resolve application network execution anomalies by optimizing the cyclic process between the application network and the execution agent, thereby ensuring the correct realization of user intentions.

[0019] In one feasible implementation, the first task includes a second subtask applied to the communication device. Before establishing a communication connection between the application network and the communication device, the method further includes: the executing agent sending the second subtask to the connecting agent; and the connecting agent configuring the second subtask into the access network device and / or the terminal device.

[0020] In an implementation, the multi-agent system further comprises a first memory, the first memory comprising information corresponding to the stored intention request and information of the first task; or the multi-agent system further comprises a second memory, the second memory comprising information corresponding to the stored first task and information of the first set of functional modules; or the multi-agent system further comprises a third memory, the third memory comprising information corresponding to the stored first network topology and resource requirement information of the functional modules of the first network topology.

[0021] In an implementation, the first memory, the second memory and the third memory are shared memories; or any two of the first memory, the second memory and the third memory are shared memories; or none of the first memory, the second memory and the third memory are shared memories.

[0022] In an implementation, the multi-agent system further comprises a second tool library, the second tool library being used to provide functional modules for assembling agents; and / or the multi-agent system further comprises a third tool library, the third tool library being used to provide tools for deploying functional modules in the first network topology.

[0023] In an implementation, the first tool library, the second tool library and the third tool library are shared tool libraries; or any two of the first tool library, the second tool library and the third tool library are shared tool libraries; or none of the first tool library, the second tool library and the third tool library are shared tool libraries.

[0024] In the embodiments of the present application, possible layout manners of the memory and the tool library, two network entities included in the multi-agent system, are provided. Independent layout of the memory and / or the tool library can guarantee independent operation of each agent and no influence on use of the network entities between the agents, thereby improving fault tolerance of the multi-agent system. Shared layout of the memory and / or the tool library can reduce implementation cost and complexity.

[0025] In a second aspect, a multi-agent system is provided, which includes an orchestration agent, an assembly agent, an execution agent, and a connection agent. The orchestration agent is configured to obtain an intent request, obtain a first task according to the intent request and a network model, and send information of the first task to the assembly agent. The first task includes at least one first subtask and at least one second subtask. The first subtask corresponds to a core network, and the second subtask corresponds to a communication device. The assembly agent is configured to receive the information of the first task from the orchestration agent, obtain a first function module set according to the at least one first subtask and a first tool library, and send information of a first network topology to the execution agent. The first function module set includes at least one function module corresponding to the at least one first subtask. The execution agent is configured to receive the information of the first network topology from the assembly agent, obtain a core network corresponding to the first network topology according to the at least one function module in the first network topology, and send connection-related information to the connection agent. The connection-related information includes information of the core network and task topology information related to the at least one second subtask in the first network topology. The connection agent is configured to receive the connection-related information, configure the at least one second subtask on the communication device according to the task topology information, and establish a communication connection between the core network and the communication device to build an application network. The communication device includes at least one of an access network device or a terminal device.

[0026] In an implementation, the orchestration agent is further configured to receive information of a second task from the orchestration agent, determine that obtaining a second function module set according to at least one first subtask in the second task and the first tool library fails, and obtain the first task according to the second task, the first response, and the network model. The assembly agent is further configured to receive the information of the second task from the orchestration agent, determine that obtaining the second function module set according to the at least one first subtask in the second task and the first tool library fails, and send a first response to the orchestration agent. The first response indicates that obtaining the second function module set fails. The orchestration agent is further configured to receive the first response.

[0027] In an implementation, the assembling agent is further configured to obtain a third set of function modules according to the at least one first subtask and the first tool library, the third set of function modules comprising at least one function module corresponding to the at least one first subtask; the assembling agent is further configured to obtain a second network topology according to the at least one function module in the third set of function modules and the at least one second subtask, and send information of the second network topology to the execution agent; the execution agent is further configured to receive the information of the second network topology from the assembling agent, and determine that a core network corresponding to the second network topology fails to be obtained according to the at least one function module in the second network topology; the execution agent is further configured to send a second response to the assembling agent, the second response indicating that the core network corresponding to the second network topology fails to be obtained; the assembling agent receives the second response from the execution agent; and the assembling agent is further configured to obtain the first set of function modules according to the second response, the at least one first subtask and the first tool library, wherein the third set of function modules and the first set of function modules comprise at least one different function module.

[0028] In an implementation, the execution agent obtains the core network corresponding to the first network topology according to the at least one function module in the first network topology, comprising: the execution agent deploys the at least one function module in the first network topology on at least one core network element, and obtains a core network composed of the at least one core network element.

[0029] In an implementation, the first network topology further comprises resource requirements of each function module in the at least one function module, and the execution agent deploys the at least one function module in the first network topology on the at least one core network element, comprising: the execution agent deploys the at least one function module in the first network topology on the at least one core network element according to the resource requirements of each function module.

[0030] In an implementation, the execution agent is configured to determine that the core network corresponding to the second network topology fails to be obtained according to the at least one function module in the second network topology, comprising: the execution agent is configured to determine that the at least one function module in the second network topology comprises a function module that fails to be deployed, and the function module fails to be deployed comprises at least one of: the core network fails to meet resource requirements of the function module; or resources of the core network fail to meet quality of service (QoS) requirements of the function module.

[0031] In an implementation, the execution agent is further configured to acquire network status information of the application network; and when the network status information indicates that the application network performs abnormally, the execution agent is further configured to adjust the application network, including at least one of the following: at least one of the number adjustment and the position adjustment of the at least one functional module; and instructing the connection agent to perform at least one of the number adjustment and the position adjustment of the at least one second subtask.

[0032] In an implementation, the first task includes a second subtask applied to the communication device, and the execution agent is further configured to send the second subtask to the connection agent before establishing the communication connection between the application network and the communication device; and the connection agent is further configured to configure the second subtask to the access network device and / or the terminal device.

[0033] In an implementation, the multi-agent system further includes a first storage configured to store the information of the intent request and the information of the first task; or the multi-agent system further includes a second storage configured to store the information of the first task and the information of the first functional module set; or the multi-agent system further includes a third storage configured to store the information of the first network topology and the resource requirement information of the functional module of the first network topology.

[0034] In an implementation, the first storage, the second storage and the third storage are shared storages; or any two of the first storage, the second storage and the third storage are shared storages; or none of the first storage, the second storage and the third storage is a shared storage.

[0035] In an implementation, the multi-agent system further includes a second tool library configured to provide the assembling agent with the functional module for assembling; and / or the multi-agent system further includes a third tool library configured to provide the functional module in the first network topology with a tool for deployment.

[0036] In an implementation, the first tool library, the second tool library and the third tool library are shared tool libraries; or any two of the first tool library, the second tool library and the third tool library are shared tool libraries; or none of the first tool library, the second tool library and the third tool library is a shared tool library.

[0037] In a third aspect, a communication apparatus is provided, which includes units or modules configured to perform any of the methods of the corresponding orchestration agent, assembling agent, execution agent or connection agent in the first aspect or any of the implementations of the first aspect.

[0038] In a fourth aspect, an embodiment of the present application provides a communication apparatus, the communication apparatus comprising at least one processor coupled with a memory; wherein the at least one processor is configured to execute computer programs or instructions stored in the memory, so that any method implemented by any of the agents in the first aspect is executed.

[0039] In a fifth aspect, an embodiment of the present application further provides a communication system, comprising the orchestration agent, the assembly agent, the execution agent and the connection agent, the access network device and the terminal device as described in the first aspect or any implementation of the first aspect.

[0040] In a sixth aspect, an embodiment of the present application provides a computer readable storage medium, wherein the computer readable storage medium stores computer instructions, when the computer instructions are executed, the computer executes any method implemented by the orchestration agent, the assembly agent, the execution agent or the connection agent as described in the first aspect.

[0041] In a seventh aspect, an embodiment of the present application provides a computer program product, comprising computer program codes, when the computer program codes are executed by a computer, the computer executes any method implemented by the orchestration agent, the assembly agent, the execution agent or the connection agent as described in the first aspect.

[0042] In an eighth aspect, an embodiment of the present application provides a chip, the chip is coupled with a memory, and is configured to read and execute program instructions in the memory, so that the apparatus in which the chip is located implements any method implemented by the orchestration agent, the assembly agent, the execution agent or the connection agent as described in the first aspect. BRIEF DESCRIPTION OF DRAWINGS

[0043] The following describes the drawings used in the embodiments of the present application.

[0044] FIG. 1 is a schematic diagram of a communication scenario provided by an embodiment of the present application.

[0045] FIG. 2A is a core network architecture based on a multi-agent system provided by an embodiment of the present application.

[0046] FIG. 2B is another core network architecture based on a multi-agent system provided by an embodiment of the present application.

[0047] FIG. 3A is a flowchart of a network management method provided by an embodiment of the present application.

[0048] FIG. 3B is a flowchart of an example of the operation of an agent system provided by an embodiment of the present application.

[0049] FIG. 4A is a schematic diagram of a cycle optimization process of a multi-agent system provided by an embodiment of the present application.

[0050] FIG. 4B-4D are flow diagrams of network management methods corresponding to the circulation optimization process according to an embodiment of the present application.

[0051] FIG. 5A-5C are network entity layout diagrams according to an embodiment of the present application.

[0052] FIG. 6 is a structural diagram of a communication device according to an embodiment of the present application.

[0053] FIG. 7 is a hardware structural block diagram of a communication device according to an embodiment of the present application. DETAILED DESCRIPTION

[0054] The technical solutions in the embodiments of the present application will be described below with reference to the accompanying drawings. In the embodiments of the present application, the terms "system" and "network" can be used interchangeably. Unless otherwise specified, " / " represents an "or" relationship between the objects before and after the slash, for example, A / B can represent A or B; in the present application, "and / or" is only a description of the relationship between the associated objects, which means that there can be three relationships, for example, A and / or B, which can represent three cases: A alone, A and B together, and B alone, where A and B can be singular or plural. In addition, in the description of the present application, unless otherwise specified, "multiple" means two or more than two. "At least one of the following" or similar expressions means any combination of these items, including any combination of single or multiple items. For example, at least one of a, b, or c can represent a, b, c, a-b, a-c, b-c, or a-b-c, where a, b, and c can be one or more. In addition, in order to clearly describe the technical solutions of the embodiments of the present application, in the embodiments of the present application, the terms "first", "second", and the like are used to distinguish the same items or similar items with the same function. Those skilled in the art can understand that the terms "first", "second", and the like do not limit the number and execution order, and the terms "first", "second", and the like do not necessarily mean different.

[0055] Reference within the specification to "one embodiment" or "some embodiments" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the application. The appearances of the phrase "in one embodiment" or "in some embodiments" in various places within specifications are not necessarily all referring to the same embodiment, however, are meant to signify that "one or more, but not all embodiments" of the application so described are contemplated to develop the application. The terms "including," "comprising," "having" and variations thereof herein are meant to encompass the items listed thereafter, but do not exclude other items from being present.

[0056] The following detailed description is provided to understand the objects, technical solutions, and advantages of the present application. It should be understood that the following is only a specific embodiment of the present application, and is not intended to limit the scope of protection of the present application. Any modification, equivalent replacement, improvement, etc. made on the basis of the technical solutions of the present application shall be included in the scope of protection of the present application.

[0057] In various embodiments of the present application, the terms and / or descriptions of different embodiments are consistent and can be referred to each other if there is no special description and logical conflict. The technical features in different embodiments can be combined to form new embodiments according to their inherent logical relationship.

[0058] The system architecture related to the embodiments of the present application is introduced as follows.

[0059] The embodiments of the present application can be applied to the fifth generation (5th generation, 5G) communication system or future mobile communication system, or other communication systems, etc., and the present application does not limit the same. The scenario of the embodiments of the present application is the background of signal transmission, which can be referred to FIG. 1. FIG. 1 is a communication scenario diagram provided by an embodiment of the present application, as shown in the figure, the communication scenario includes terminal(s), base station(s) and core network. The embodiments of the present application are applicable to homogeneous network and heterogeneous network scenarios, and there is no limitation to transmission points, which can be multi-point cooperative transmission between macro base stations and macro base stations, micro base stations and micro base stations, and macro base stations and micro base stations. It is applicable to frequency division duplexing (frequency division duplexing, FDD) or time division duplexing (time division duplexing, TDD) system. The embodiments of the present application are applicable to low frequency scenarios (sub 6GHz) and high frequency scenarios (above 6GHz), terahertz, optical communication, etc.

[0060] The terminal involved in the embodiments of the present application can also be referred to as a terminal device, a UE, etc. The terminal device can be an entity for receiving or transmitting signals on the user side, such as a mobile phone. The terminal device can be used to connect people, things and machines. The terminal device can communicate with one or more core networks through a network device. The terminal device includes a handheld device with a wireless connection function, another processing device connected to a wireless modem, or a vehicle-mounted device, etc. The terminal device can be a portable, pocket-sized, handheld, computer-built-in or vehicle-mounted mobile device. The terminal device can be widely used in various scenarios, such as cellular communication, D2D, V2X, point-to-point (P2P), machine-to-machine (M2M), machine type communication (MTC), internet of things (IoT), virtual reality (VR), augmented reality (AR), industrial control, autonomous driving, remote medical treatment, smart power grid, smart furniture, smart office, smart wear, smart transportation, smart city, unmanned aerial vehicle, robot, remote sensing, passive sensing, positioning, navigation, autonomous delivery and mobile, etc.Some examples of the terminal device 120 are: a user equipment (UE) of a 3GPP standard, a fixed device, a mobile device, a handheld device, a wearable device, a cellular phone, a smart phone, a session initiated protocol (SIP) phone, a notebook, a personal computer, a smart book, a vehicle, a satellite, a global positioning system (GPS) device, a target tracking device, a drone, a helicopter, an aircraft, a ship, a remote control device, a smart home device, an industrial device, a personal communication service (PCS) phone, a wireless local loop (WLL) station, a personal digital assistant (PDA), a wireless webcam, a tablet, a palm computer, a mobile internet device (MID), a wearable device such as a smart watch, a VR device, an AR device, a wireless terminal in industrial control, a terminal in Internet of Vehicles, a wireless terminal in self-driving, a wireless terminal in smart grid, a wireless terminal in transportation safety, a wireless terminal in smart city such as a smart refueling device, a terminal device on a high-speed rail, and a wireless terminal in smart home such as a smart sound box, a smart coffee machine, a smart printer, etc. The terminal device 120 can be a wireless device in the above various scenarios or an apparatus for being arranged in a wireless device, e.g., a communication module, a modem, or a chip in the above devices. The terminal device can also be referred to as a terminal, a terminal device, a user equipment (UE), a mobile station (MS), a mobile terminal (MT), etc. The terminal device can also be a terminal device in a future wireless communication system. The terminal device can be used in a dedicated network device or a general-purpose device. The embodiments of the present application do not limit the specific technology and specific device form of the terminal device.

[0061] Optionally, the terminal device can contain part of the functions of a base station. For example, a UE can act as a scheduling entity, which provides a sidelink signal between UEs in V2X, D2D, or P2P, etc. As shown in FIG. 1, a mobile phone and a drone communicate with each other using a sidelink signal.

[0062] In this application, the communication device for realizing the function of the terminal device can be a terminal device, a terminal device with part of the function of the terminal device, or a device capable of supporting the function of the terminal device, such as a chip system, which can be installed in the terminal device or matched with the terminal device. In this application, the chip system can be composed of a chip or include a chip and other discrete devices. In the technical solutions provided in this application, the communication device is taken as an example for description.

[0063] The base station (BS) involved in the embodiments of the present application can also be referred to as a radio access network device, an access network network element, a radio access network (RAN) node (or device, or network element), an access point (AP), a network device, a small tower, etc. The base station can be variously named or replaced by the following names in a broad sense, such as a radio access network (RAN) node, a NodeB, an evolved NodeB (eNB), a next generation NodeB (gNB), an access network device in an open radio access network (O-RAN), a relay station, an access point, a transmitting and receiving point (TRP), a transmitting point (TP), a master eNB (MeNB), a secondary eNB (SeNB), a multi-standard radio (MSR) node, a home base station, a network controller, an access node, a wireless node, an access point (AP), a transmission node, a transceiver node, a building baseband unit (BBU), a remote radio unit (RRU), an active antenna unit (AAU), a remote radio head (RRH), a centralized unit (CU), a distributed unit (DU), a radio unit (RU), a centralized unit control plane (CU-CP) node, a centralized unit user plane (CU-UP) node, a positioning node, etc. The base station can be a macro base station, a micro base station, a relay node, a donor node, or the like, or a combination thereof. The network device can also refer to a communication module, a modem, or a chip for being disposed in the foregoing devices or apparatuses. The network device can also be a mobile switching center and a device assuming a base station function in device-to-device (D2D), vehicle-to-everything (V2X), machine-to-machine (M2M) communication, a network side device in a future communication system, etc.The network device can support networks of the same or different access technologies. Embodiments of the present application do not limit the specific technology and specific device form adopted by the network device.

[0064] In some deployments, the network device mentioned in the embodiments of the present application can be a device including a CU, or a DU, or a device including a CU and a DU, or a control plane CU node (central unit-control plane (CU-CP)) and a user plane CU node (central unit-user plane (CU-UP)) and a DU node. For example, the network device can include a gNB-CU-CP, a gNB-CU-UP and a gNB-DU.

[0065] In some deployments, a plurality of RAN nodes cooperate to assist a terminal to implement wireless access, and different RAN nodes respectively implement part of the functions of a base station. For example, the RAN node can be a CU, a DU, a CU-CP, a CU-UP, or an RU, etc. The CU and the DU can be separately arranged, or can be included in the same network element, such as a BBU. The RU can be included in a radio frequency device or a radio frequency unit, such as an RRU, an AAU or an RRH.

[0066] The network device is responsible for all functions related to the air interface, including: (1) a radio link maintenance function, maintaining a radio link with a terminal, and being responsible for protocol conversion of radio link data and IP data quality monitoring; (2) a radio resource management function, including establishment and release of a radio link, scheduling and allocation of radio resources, etc.; (3) part of the mobility management function, including configuring the terminal to perform measurement, evaluating the quality of the radio link of the terminal, and deciding the handover of the terminal between cells, etc.

[0067] In the present application, the communication device for implementing the access network function as described above can be an access network device, or a network device having part of the function of the access network, or a device capable of supporting the implementation of the access network function, such as a chip system, a hardware circuit, a software module, or a hardware circuit plus a software module, which can be installed in the access network device or used in conjunction with the access network device. In the method of the present application, the communication device for implementing the function of the access network device is taken as an example for description.

[0068] The core network (CN) involved in the embodiments of the present application mainly provides user connection, management of users, and completion of bearer for services, and provides an interface to external networks as a bearer network.

[0069] It should be understood that the number and type of devices in the communication system shown in FIG. 1 are merely illustrative, and the present application is not limited thereto, and in actual applications, more terminal devices, more access network devices, and other network elements, such as network elements for implementing artificial intelligence functions, can also be included in the communication system.

[0070] It can be understood that all or part of the functions implemented by one or more of the terminal device, the access network device, the core network device, or the network element for implementing the artificial intelligence function can be virtualized, that is, implemented by one or more of a special processor or a general processor and a corresponding software module. Among them, the terminal device and the access network device involve the interface of air interface transmission, and the transceiving function of the interface can be implemented by hardware. The core network device, such as the operation administration and maintenance (OAM) network element, can be virtualized. Optionally, one or more functions of the virtualized terminal device, access network device, core network device, or network element for implementing the artificial intelligence function can be implemented by a cloud device, such as a cloud device in an over the top (OTT) system.

[0071] The prior art of the embodiments of the present application is introduced below.

[0072] Future network technologies need to support various business scenarios and use cases, including immersive communication, intelligent industry, and digital medical treatment, etc. These new businesses have various differences in performance requirements, which will greatly increase the complexity of network functions and the difficulty of management and configuration. In order to better serve these new businesses, the future network needs to have strong on-demand customization capabilities, and needs to be able to arrange and configure various functions and resources in the network in a more flexible and dynamic way. This is a great challenge for traditional networks that define workflows through standards, and the introduction of AI can help solve this problem.

[0073] The idea of integrating communication networks with AI technology is proposed, which can be simply referred to as network AI. Network AI is to provide a complete AI environment and AI services in the network through a unified architecture design. This will bring great changes to the traditional network architecture, from the traditional connection-oriented service-oriented session-centric architecture to the AI-oriented new service-oriented task-centric architecture. For example, by introducing a new dimension of "task" into the future communication network architecture, the QoS of network AI services can be better controlled and guaranteed at the network execution level. This also makes it possible to have a measurable dimension at the network operation level, which can support the new generation of network architecture, such as the new business model of future communication networks.

[0074] Large models exhibit strong intent understanding and intelligence emerging capabilities, and their application scenarios are more extensive. For example, ChatGpt can accurately identify and understand user voice and text information, and provide more intelligent services. The integration of communication networks and AI technology is more likely to be the integration of communication networks and large models. Previous AI was trained from scratch by collecting a large amount of network data, and it was necessary to focus on how to design the model structure for specific tasks, how the network collected the required data, and how the network provided computing power and algorithm support for AI model training and inference. However, the application paradigm of large models is based on pre-training of basic models, and various strategies such as prompts, fine-tuning, and vector libraries are used to adapt to various tasks. The huge amount of computing power and communication demand brought by the large number of parameters of large models also brings new challenges to their application in networks.

[0075] Currently, in order to realize the application of large models in actual scenarios, AI agents are proposed. AI agents refer to functional entities that provide AI services for applications. The implementation of AI agents requires a large model (or AI model) as the core, combined with a series of peripheral functions to support the system engineering. Therefore, future networks also need to be redesigned based on the idea of AI agents, so that the network can be an AI-enabled infrastructure that provides intelligent services for various applications. Among them, the functions that AI agents need to implement include:

[0076] (1) Task decomposition of the received user intent.

[0077] The user's intent can be natural language and can correspond to a large task target. AI agents can understand user intent and decompose large task targets into multiple easily managed subtasks, thereby efficiently processing complex tasks.

[0078] (2) Call memory storage.

[0079] AI agents can call memory storage functions, including short-term memory storage and long-term memory storage. Short-term memory storage is used for large model learning, that is, calling stored relevant context, knowledge, historical data, and other existing tool information, combining with the prompt requirements of the large model, generating the input of the large model. The large model outputs the decomposition task. Long-term memory storage is used for long-term storage of information.

[0080] (3) Application of tool library.

[0081] The AI agent can call external application interfaces through the tool library to obtain additional information missing in the large model weight, including current information, code execution capability, access to proprietary information sources, etc. In addition, the function modules in the tool library can be called to form network functions and application functions corresponding to the aforementioned decomposed subtasks, forming an application network topology.

[0082] (4) Execution of the application network.

[0083] The formed application network topology is instantiated (application network instance is generated), and the network functions and application functions therein are configured to the core network. At the same time, the base station or UE that needs to be accessed is configured to execute the application network instance. In addition, the AI agent also needs to periodically manage and control the generated application network instance, including modification and deletion, etc.

[0084] The above describes the functions or services that the AI agent needs to provide. However, how to design the network architecture, including deploying function entities in the network architecture, establishing communication connections between the function entities and the network elements in the network architecture, etc., to realize these functions or services, is a problem to be solved.

[0085] Based on the above problems, the embodiments of the present application provide a core network architecture based on a multi-agent system, which includes a multi-agent system and other core network elements connected to the multi-agent system, to realize the functions or services of the AI agent. Referring to FIG. 2A, the multi-agent system includes:

[0086] A planning agent is responsible for task planning.

[0087] The agent described in the embodiments of the present application refers to an entity, network element, or chip in an entity device that has certain or some functions, or can be used to realize certain or some functions. The agents involved below are applicable to the description, and will not be described in detail.

[0088] Exemplarily, the planning agent serves as an interface of the application network service and can receive user intent so that the multi-agent system provides customized application network services based on the user intent. That is, the planning agent decomposes the received user intent into subtasks using a network model and sends them to the assembly agent. Optionally, the planning agent can also recycle the last resources after the task execution is completed.

[0089] Optionally, the subtasks include a first subtask and a second subtask, wherein the first subtask is used to build the part of the application network related to the core network, and the second subtask is used to build the part of the application network related to the communication device.

[0090] Optionally, the communication device comprises one or more access network devices; or the communication device further comprises one or more terminal devices.

[0091] An assembling agent is configured to be responsible for function assembling.

[0092] The assembling agent is configured to obtain a function module corresponding to the first subtask. That is, the assembling agent invokes a tool library based on the result of task orchestration, matches a corresponding function module for each first subtask, including network function and application function. The tool library comprises various function modules.

[0093] Optionally, the tool library can be a shared tool library of the multi-intelligent system. As shown in FIG. 2A, the tool library can be in communication connection with all agents in the multi-intelligent system.

[0094] Optionally, the tool library can be individually corresponding to the orchestration agent. The tool library is connected with the orchestration agent.

[0095] The assembling agent is configured to obtain a function module corresponding to the first subtask. That is, the assembling agent invokes a tool library based on the result of task orchestration, matches a corresponding function module for each first subtask, including network function and application function. The tool library comprises various function modules.

[0096] An action agent is configured to be responsible for management and control of the application network (instance).

[0097] The assembling agent is configured to obtain a function module corresponding to the first subtask. That is, the assembling agent invokes a tool library based on the result of task orchestration, matches a corresponding function module for each first subtask, including network function and application function. The tool library comprises various function modules.

[0098] Optionally, the assembling agent is connected with the application network (as shown in FIG. 2A) and is configured to send the application network topology information to the network elements or devices in the application network. The application network topology information comprises interfaces of the function modules, execution sequence, etc.

[0099] The action agent is configured to control execution of the application network, for example, sending execution instructions to the network elements or devices in the application network, etc.

[0100] Exemplarily, the execution agent is further configured to send information of the core network and task topology information related to the second subtask in the application network topology to the connecting agent, so that the connecting agent establishes a connection between a network element in the core network and a communication device corresponding to the second subtask, and completes construction of the final application network.

[0101] Exemplarily, the execution agent is further configured to monitor network status and timely optimize and adjust the application network instance.

[0102] The connecting agent is configured to establish a connection between the application network and the communication device.

[0103] Exemplarily, the connecting agent obtains the information of the core network and the task topology information of the second subtask from the execution agent, and establishes a connection between a network element in the core network and a communication device corresponding to the second subtask.

[0104] Optionally, the information of the core network includes an identifier of the core network, or a core network element included in the core network, and an execution order of an interface between the network elements and a functional module on the network element.

[0105] Optionally, the task topology information of the second subtask includes an access network device (base station) or a terminal device (UE) to which the second subtask needs to be configured, and a connection interface, an execution order, and a connection relationship between the second subtask and a network element in the core network between the second subtasks.

[0106] The connecting agent is connected with the UE, the base station, and an application gateway (AGW), and is configured to establish a connection between the UE, the base station, and the core network, to call a RAN resource to establish data routing between the UE and the application network, and to form the application network. The AGW is configured to provide an application network data anchor or proxy, and the UE and the base station are configured to provide data support for the application network, and thus are part of the application network.

[0107] The four agents are connected with each other and are configured to realize the functions of the AI agent.

[0108] Optionally, the application network is deployed only on the core network element.

[0109] Optionally, the subtasks decomposed from the user intention include at least one first subtask.

[0110] Optionally, the assembling agent in the multi-agent system obtains an application network topology according to at least one function module corresponding to at least one first subtask, and the application network topology only includes topology information between the function modules.

[0111] Optionally, the performing agent completes deployment of at least one function module in the application network topology, and after obtaining the core network, the creation of the application network is completed.

[0112] Optionally, the multi-agent system does not include a connecting agent. For example, after obtaining the core network, the creation of the application network is completed, and the communication device is not needed to complete the application network, that is, the connecting agent is not needed to be set to connect the core network element and the communication device.

[0113] Referring to FIG. 2B, FIG. 2B is another core network architecture based on a multi-agent system provided by an embodiment of the present application. In the network architecture, an orchestrating agent, an assembling agent and a performing agent are included, and the connection mode of the three agents and other core network functions can be referred to FIG. 2A. Herein, no further description is given.

[0114] Optionally, in the above core network architecture, the agent in each multi-agent system can include an independent memory, that is, each agent is connected with an independent memory.

[0115] The tool library and the memory will be described in subsequent embodiments, and no further description is given herein.

[0116] The specific workflow of the multi-agent system shown in FIG. 2A is described below.

[0117] Referring to FIG. 3A, FIG. 3A is a flowchart of a network management method provided by an embodiment of the present application, which can be applied to the multi-agent system shown in FIG. 2A. As shown in FIG. 3A, the method includes the following steps:

[0118] 101. The orchestrating agent obtains an intent request, obtains a first task according to the intent request and a network model, and sends information of the first task to the assembling agent.

[0119] The first task includes at least one first subtask and at least one second subtask, the first subtask corresponds to the core network, and the second subtask corresponds to the communication device.

[0120] Optionally, the first subtask corresponds to the core network, which means that the first subtask is executed or implemented by the core network element.

[0121] Optionally, the second subtask corresponds to the communication device, which means that the second subtask is executed or implemented by the communication device.

[0122] An intent request refers to a user sending intent-related information for requesting to complete one or more tasks or achieve one or more goals. A network model is used to understand and process the intent request, and decompose the intent request into one or more subtasks, which are referred to as first tasks in embodiments of the present application.

[0123] In embodiments of the present application, the orchestration agent, the assembly agent, the execution agent, and the connection agent in the multi-agent system are communicatively connected. After obtaining the first task, the orchestration agent can send the information of the first task to the assembly agent through the communication connection with the assembly agent. Correspondingly, the assembly agent receives the information of the first task and obtains the first task.

[0124] Optionally, the information of the first task includes at least one of a name or an identifier of the first task, and / or includes at least one of a name or an identifier of a subtask in the first task.

[0125] The method of embodiments of the present application will be described in combination with a specific example of FIG. 3B. FIG. 3B is a flowchart of an example of working of an agent system according to embodiments of the present application. As shown in FIG. 3B, the intent request of the user can be natural language, and the specific content can be, for example, “Please count the total number of people and vehicles at the No. 1 gate of Shanghai Century Park during 8:00-17:00”. After receiving the intent request, the orchestration agent performs task orchestration. Specifically, the intent request is input into the network model, and other content, such as knowledge, historical data, prompt information, etc., is input into the network model, so that the network model outputs the first task. The knowledge can be, for example, a list of subtasks; the historical data can be, for example, a correspondence between historical intent requests and subtasks; and the prompt information can be, for example, information about the purpose of the task, such as counting the number of people and vehicles. As shown in FIG. 3B, the intent request described above is decomposed into a total of five subtasks, 1-5, after task orchestration, including 1. extracting the time and location input by the user; 2. converting the location input by the user into latitude and longitude, and querying the base station according to the latitude and longitude; 3. collecting perception data from the base station, etc.

[0126] Among them, subtasks 1, 2, 4, and 5 are subtasks executed or implemented by the core network element, and thus are first subtasks.

[0127] Task 3 requires the participation of the RAN device, and thus is a second subtask.

[0128] In an optional case, a subtask decomposed from the user intent can need to be executed by both the core network element and the terminal device, and thus the same subtask can be both a first subtask and a second subtask.

[0129] 102. Assemble the intelligent agent to obtain a first set of functional modules based on at least one first subtask and a first tool library.

[0130] The first functional module set includes at least one functional module corresponding to at least one first subtask.

[0131] Optionally, the first tool library stores information related to multiple functional modules, including the names or identifiers of the functional modules, or it may also store descriptions of each functional module. Matching a subtask with a functional module can be done by keyword matching between the names of the subtask and the functional module, or it may also include keyword matching between the descriptions of the subtask and the functional module. For example, as shown in Figure 3B, for the subtask "Perform object detection," the functional module "Object Detection" can be matched and obtained from the first tool library.

[0132] Optionally, the first tool library is an (independent) tool library corresponding to the assembly agent.

[0133] The assembly agent obtains a first set of functional modules based on the first subtask and the first tool library, including: the assembly agent matching the information of each first subtask with the relevant information of the functional modules in the first tool library to obtain the functional modules corresponding to each subtask, and forming the first set of functional modules.

[0134] Optionally, there is a one-to-one correspondence between the first subtask and the functional module, that is, one first subtask matches one functional module.

[0135] Optionally, the first subtask and the functional module have a one-to-many relationship, that is, one first subtask can be matched with multiple functional modules.

[0136] Optionally, the relationship between the first subtask and the functional module is many-to-many, meaning that the functional modules matched by multiple first subtasks have overlapping implementation processes.

[0137] 103. The assembling agent obtains the first network topology based on at least one functional module and at least one second subtask from the first functional module set, and sends the information of the first network topology to the executing agent. Correspondingly, the executing agent receives the information of the first network topology.

[0138] After assembling the intelligent agent to obtain the first set of functional modules, it further assembles one or more functional modules and one or more second subtasks, establishes interconnection relationships between the functional modules and the second subtasks, connects their interfaces, determines the execution order of each functional module and the second subtask, and forms the first network topology.

[0139] Optionally, the first network topology further comprises resource requirements of each function module. That is, when the assembling agent assembles the function modules, the resource requirements of the modules are also determined. For example, the resource requirements of the function module of "object detection" include that the CPU requirement is greater than 1 GHz and the memory requirement is less than 10 nanoseconds (ns).

[0140] Optionally, the information of the first network topology comprises at least one of a name or an identifier of the first network topology, and / or at least one of a name or an identifier of the function module in the first network topology.

[0141] Optionally, the information of the first network topology further comprises at least one of an interconnection relationship, a connection interface, or an execution sequence of the function module and the second subtask that constitute the first network topology.

[0142] Optionally, the information of the first network topology further comprises resource requirements of at least one of the function module and the second subtask that constitute the first network topology.

[0143] 104. The execution agent obtains, according to at least one function module in the first network topology, a core network corresponding to the first network topology.

[0144] The assembling agent sends the first network topology to the execution agent, and the execution agent deploys the function modules in the first network topology according to the corresponding interface connection relationship, execution sequence, etc.

[0145] For example, deploying the function modules is to deploy the function modules to the core network, to determine that a certain network element in the core network executes the function module, or to deploy the function modules to a new network element, and then connect the new network element to the core network to execute the function module. For example, the function module of "BS sensing" in FIG. 3B can be deployed in an access and mobility management function (AMF) network element.

[0146] Optionally, the first network topology further comprises resource requirements of each function module, and deploying the function modules in the first network topology comprises: deploying each function module in the first network topology according to the resource requirements of each function module.

[0147] As described above, the first network topology can further comprise resource requirements of each function module. Therefore, when deploying the function modules, the resource requirements of each function module also need to be considered (corresponding to the resource scheduling process in FIG. 3B). For example, the resource requirements of the function module of "BS sensing" are that the CPU requirement is greater than 1 GHz, and then it needs to be determined whether the AMF network element in the core network can meet the CPU requirement.

[0148] Optionally, the function module deployment failure comprises at least one of the following: the core network does not meet the resource requirement of the function module; or the resource of the core network does not meet the quality of service (QoS) requirement of the function module.

[0149] In the above process, each function module is deployed according to its own resource requirement. This means that the core network element to which the function module is deployed may or may not be able to meet the resource requirement. If it can meet the requirement, the deployment is successful; if it cannot meet the requirement, it may mean that the function module deployment fails.

[0150] Or in some cases, the core network element can meet the resource requirement of the function module, and the function module deployment is successful. However, during the execution of the function module, the QoS requirement of the function module cannot be met. For example, the function module "BS awareness" is deployed on the AMF element, but during the execution of the function module, it is found that the QoS requirement of less than 1 ns delay cannot be met. Therefore, the function module is also considered to be deployed unsuccessfully. After the function module deployment fails, there can be corresponding solutions, including re-matching to obtain a new function module, re-dividing sub-tasks to obtain a new function module, and the like, which will be described in subsequent embodiments.

[0151] In this embodiment, the way to determine that the function module deployment fails includes that the core network does not meet the resource requirement of the function module; or the resource of the core network does not meet the quality of service (QoS) requirement of the function module. The former can determine that the function module deployment fails in the deployment stage, and the latter can determine that the function module deployment fails in the execution stage. The determination can meet the deployment of the function module in each stage.

[0152] After the execution agent completes the deployment of all the function modules, the creation of the core network is completed, and the core network is used to implement the first sub-task part.

[0153] 105、The execution agent sends connection-related information to the connection agent, the connection-related information including information of the core network and task topology information related to at least one second sub-task in the first network topology. Correspondingly, the connection agent receives the connection-related information.

[0154] 106、The connection agent configures the at least one second sub-task to the communication device according to the task topology information.

[0155] 107、The connection agent establishes a communication connection between the core network and the communication device to build an application network.

[0156] The communication device includes at least one of an access network device or a user equipment.

[0157] The use of the application network generally includes providing network services for users, or implementing network awareness, etc. The participation of at least one of the communication devices, i.e., the UE and the RAN device (generally, the RAN is also included in the scenario of communication between the CN and the UE, because the CN does not communicate directly with the UE) is generally required in these uses. Therefore, in the process of constructing the application network, the communication connection between the core network and the communication device is also constructed, so that the communication device participates in the implementation of the functions of the application network. For example, the aforementioned function module "BS awareness" requires the core network to send a message to the RAN device, and the RAN device feeds back the awareness data to the core network to complete the execution of the function module.

[0158] For example, in the first task, the subtask executed or implemented by the communication device is the second subtask. Optionally, the task topology information related to the second subtask can include: a. information of the second subtask, such as the name or identifier of the second subtask, etc.; b. topology information of the second subtask, such as the connection relationship between the second subtasks, the execution order, the connection relationship between the second subtask and at least one function module in the core network (or at least one network element to which the function module is deployed), the execution order, etc.; c. configuration information of the second subtask. That is, the second subtask needs to be configured to which communication device, the amount or quantity configured on the communication device, etc.

[0159] For example, after the connection agent obtains the task topology information related to the second subtask, the second subtask is configured to the RAN device or the UE, or the second subtask needs to be deployed in both devices, including deploying the same second subtask (in different steps) or deploying different second subtasks in both. This process can correspond to the subtask configuration step in FIG. 3B. Among the five subtasks arranged by the intention request, subtask 3 is to collect awareness data from the base station, which requires the participation of the RAN device to provide the awareness data, so the steps related to the RAN in this subtask are configured to the RAN device.

[0160] Then the connection device establishes the communication connection between the core network and the communication device to complete the construction of the application network, which can be targeted to implement the user intention and complete the execution of the first task.

[0161] It can be seen that the embodiments of the present application construct a multi-agent system. The arrangement agent, the assembly agent, the execution agent, and the connection agent in the multi-agent system are respectively used for implementing task arrangement, matching and assembly of tools, deployment and connection of the core network and the communication device, etc., and the customized application network service is completed through the interaction between the multi-agents. The method process realizes the function service of the agent, and the implementation mode of the multi-agent reduces the implementation difficulty of the agent and improves the flexibility and fault tolerance of the system.

[0162] The above embodiments describe the case that the assembly agent normally obtains the first function module set, and the function modules in the first function module set are normally deployed and executed. In fact, various abnormal situations can occur in these processes. Therefore, the embodiments of the present application also provide a circulation optimization process of a multi-agent system, which can be specifically referred to FIG. 4A. As shown in FIG. 4A, the circulation optimization process of the multi-agent system includes:

[0163] Cycle ①: the cycle between the scheduling agent and the assembly agent. It can also be referred to as outer cycle. It is used to handle the case that the sub-tasks decomposed by the scheduling agent cannot be completely matched with the obtained function modules.

[0164] Cycle ②: the cycle between the assembly agent and the execution agent. It can also be referred to as middle cycle. It is used to handle the case that the function modules in the network topology assembled by the assembly agent cannot be completely deployed.

[0165] Cycle ③: the cycle between the execution agent and the application network. It can also be referred to as outer cycle. It is used to handle the case that the function modules in the application network cannot be normally or efficiently executed.

[0166] The three-layer cycle is described in detail below in combination with the drawings.

[0167] FIG. 4B is a flowchart of another network management method provided by the embodiments of the present application. As shown in FIG. 4B, on the basis of the network management method disclosed in the foregoing FIG. 3A, in combination with the above cycle ①, before the scheduling agent obtains the first task, the method can further include the following steps:

[0168] 201: The scheduling agent obtains the intention request, obtains the second task according to the intention request and the network model, and sends the information of the second task to the assembly agent. The second task also includes at least one first sub-task and at least one second sub-task, and the second task includes at least one different first sub-task from the first task. Correspondingly, the assembly agent receives the information of the second task from the scheduling agent.

[0169] 202: The assembly agent determines that the acquisition of the second function module set according to the at least one first sub-task in the second task and the first tool library fails.

[0170] Optionally, if the acquisition of the corresponding function module of any one of the first sub-tasks in the second task fails, it is determined that the acquisition of the second function module set fails.

[0171] Optionally, if the acquisition of the corresponding function module of more than a preset number of first sub-tasks in the second task fails, it is determined that the acquisition of the second function module set fails. The preset number is less than or equal to the total number of the first sub-tasks included in the second task.

[0172] 203、the assembling agent sends a first response to the orchestration agent, the first response indicating that the obtaining of the second set of function modules fails; correspondingly, the orchestration agent receives the first response.

[0173] Optionally, the first response can indicate the failure of obtaining the set of function modules by indicating one of the following information:

[0174] a. the first subtask in the second task in which the matching of the function module fails.

[0175] In this method, it can not be indicated which first subtask matching fails (which can be referred to as full matching failure).

[0176] b. the specific first subtask in the second task in which the matching of the function module fails.

[0177] For example, it indicates that the first subtask 1 matching function module fails (which can be referred to as individual matching failure).

[0178] 204 (101), the orchestration agent obtains the first task according to the intent request and the network model, including: obtaining the first task according to the second task, the first response and the network model.

[0179] Among them, the second task and the first task include different first subtasks, that is, at least one first subtask is different.

[0180] Optionally, step 204 can replace step 101 in the foregoing embodiment related to FIG. 3A. That is, the intent request is characterized by the second task and the first response.

[0181] For the above two kinds of first response, that is, indicating full matching failure or indicating individual matching failure, the former indicating method, when the network model processes according to the first task and the first response, has greater freedom, and when outputting the second task, due to the lack of specific direction, the output efficiency is also higher. For the latter indicating method, the network model can obtain more accurate output results.

[0182] Before the first subtask in the second task matching the function module fails, the above steps 201-204 can be repeatedly executed multiple times. After obtaining the first task, the subsequent process can be executed. The subsequent process is the same as the method in the foregoing FIG. 3A, and specific reference can be made to the related description of the foregoing steps 101-107.

[0183] As can be seen, in the embodiments of the present application, when determining that the subtask matching function module fails, the cycle optimization process between the assembling agent and the orchestration agent is adopted to realize the complete matching of the subtask to the function module. The normal implementation of the user intent is guaranteed.

[0184] Fig. 4C is a flow diagram of another network management method according to an embodiment of the present application. As shown in Fig. 4C, on the basis of the network management method disclosed in Fig. 3A, in combination with the above-mentioned cycle ②, before the assembling agent obtains the first function module set, the method can further include the following steps:

[0185] 301. The assembling agent obtains a third function module set according to the at least one first subtask and the first tool library, the third function module set including at least one function module corresponding to the at least one first subtask.

[0186] 302. The assembling agent obtains a second network topology according to at least one function module in the third function module set and the at least one second subtask, and sends information of the second network topology to the executing agent.

[0187] 303. The executing agent receives the information of the second network topology from the assembling agent, and determines that obtaining the second network topology according to at least one function module in the second network topology fails.

[0188] For example, the obtaining of the second network topology corresponding to the core network fails, i.e., the second network topology includes at least one function module that fails to be deployed. Alternatively, the problem can be overcome by re-obtaining the function module set corresponding to the at least one first subtask.

[0189] 304. The executing agent sends a second response to the assembling agent, the second response indicating that the obtaining of the second network topology corresponding to the core network fails; correspondingly, the assembling agent receives the second response from the executing agent.

[0190] Similar to the above-mentioned embodiment, the second response can only indicate that the first network topology includes a function module that fails to be deployed (which can be referred to as a total deployment failure) without indicating which specific function module fails to be deployed. Alternatively, the second response can indicate the specific function module in the first network that fails to be deployed, for example, the first function module (which can be referred to as an individual matching failure).

[0191] 305(102). The assembling agent obtains the first function module set according to the at least one first subtask and the first tool library, including: obtaining the first function module set according to the at least one first subtask, the first tool library and the second response.

[0192] The third function module set and the first function module set include at least one different function module.

[0193] Alternatively, the step 305 can replace the step 102 in the above-mentioned embodiment of Fig. 3A. That is, the second response is used as additional information together with the at least one first subtask and the first tool library to obtain the first function module set.

[0194] In some cases, the same first subtask can be implemented by different functional modules, or can be implemented by different functional module combinations. Therefore, at least one functional module in the first functional module set can be re-matched from the first tool library according to the second response and the information of the first task, and the first functional module set can be deployed in its entirety and obtain the corresponding core network.

[0195] Before obtaining the core network corresponding to the second network topology fails, the above steps 301-305 can be repeatedly executed multiple times. After obtaining the first functional module set, the subsequent process can be executed. The subsequent process is the same as the method in the foregoing FIG. 3A, and specific reference can be made to the related description of the foregoing steps 103-107.

[0196] Or the embodiment of the present application can also combine the method of the foregoing FIG. 4B to perform a larger cycle. In the method in the foregoing FIG. 3A, before step 103, the following steps (not shown in the figure) can also be specifically included:

[0197] 3041、The assembling agent obtains a third functional module set according to at least one first subtask and the first tool library, and the third functional module set includes at least one functional module corresponding to the at least one first subtask.

[0198] 3042、The assembling agent obtains a second network topology according to at least one functional module in the third functional module set and at least one second subtask, and sends information of the second network topology to the executing agent.

[0199] 3043、The executing agent receives the information of the second network topology from the assembling agent, and determines that obtaining the core network corresponding to the second network topology fails according to at least one functional module in the second network topology.

[0200] 3044、The executing agent sends a third response to the orchestrating agent, and the third response indicates that obtaining the core network corresponding to the second network topology fails; correspondingly, the orchestrating agent receives the third response.

[0201] Exemplarily, when the executing agent determines that obtaining the core network corresponding to the second network topology fails, the executing agent sends the third response to the orchestrating agent to re-orchestrate the obtained task, and the newly orchestrated task can match a new functional module set, instead of matching a new functional module set according to the original task.

[0202] 3045、The orchestrating agent obtains a first task according to the intention request and the network model, including: obtaining the first task according to the third task, the third response, and the network model.

[0203] The network model can obtain a third task according to the original orchestration, a function module with a deployment failure indicated by the third response, and a new first task obtained by combining the network model.

[0204] It can be seen that in the embodiments of the present application, when the deployment of the function module fails, the complete deployment of the function module is implemented by assembling the cyclic optimization process between the intelligent agent and the execution intelligent agent. The complete implementation of the user intention is ensured. Further, assuming that the deployment of the function module fails multiple times, the complete deployment of the function module can be implemented by assembling the cyclic optimization process among the orchestration intelligent agent, the assembly intelligent agent and the execution intelligent agent. The complete implementation of the user intention is further ensured.

[0205] FIG. 4D is a flow diagram of another network management method provided by the embodiments of the present application. As shown in FIG. 4D, in the network management method disclosed in the foregoing FIG. 3A, in combination with the above-mentioned cycle ③, the following steps can be further included:

[0206] 401. The execution intelligent agent obtains network state information of the application network, which includes information indicating abnormal execution of the application network.

[0207] 402. The execution intelligent agent adjusts the application network.

[0208] Exemplarily, the adjustment of the application network includes at least one of the following:

[0209] at least one of the number adjustment and the position adjustment of the at least one function module;

[0210] at least one of the number adjustment and the position adjustment of the at least one second subtask.

[0211] Exemplarily, the network state information includes network delay, transmission packet loss rate, channel state, etc. If the network state information indicates that the network delay is greater than the required delay, the packet loss rate is greater than the required packet loss rate, and the channel state is deteriorated, it indicates that the application network is abnormally executed.

[0212] The abnormal execution of the application network can be caused by the abnormal execution of the function module on the core network, such as insufficient CPU resources of the core network element. It can also be caused by the abnormal execution of the second subtask on the communication device, such as low UE power and poor communication quality. If it is determined that the former causes the abnormality, at least one function module in the network topology can be adjusted, and if it is the latter that causes the abnormality, the second subtask on the communication device can be adjusted.

[0213] The specific adjustment mode includes quantity adjustment or position adjustment, or both. The quantity adjustment includes reducing the number of processes of the function module or the second subtask, or even directly suspending the execution of a certain function module or second subtask. The position adjustment includes migrating the function module deployed on the A network element to the B network element, which can be the same type of network element as the A network element or a different type of network element. Or migrating the second subtask executed on UE1 to UE2 for execution, assuming that UE2 is not connected to the application network, a connection between the core network and UE2 also needs to be established. UE1 and UE2 can be UEs in close physical proximity.

[0214] As can be seen, in the embodiments of the present application, for the case of application network exception execution, the number of second subtasks on the communication device or the function modules in the core network can be adjusted, or the position adjustment can be performed, the former guarantees the normal use of the communication device or the core network element, and the latter guarantees the normal execution of the application network. Overall, the reliability of the application network implementation is improved.

[0215] As can be seen, in the embodiments of the present application, the problem of discovering and solving the application network execution exception through the cyclic optimization process between the application network and the execution agent is disclosed, which guarantees the correct implementation of the user's intention.

[0216] The cycle can also be combined with the first two cycles respectively, or simultaneously. For example, simultaneous combination, that is, the subtask is matched to the function module through cycle ①, and the function module is successfully deployed through cycle ②. Finally, through cycle ③, it is guaranteed that the function module is normally executed. Overall, it guarantees the correct and complete implementation of the user's intention.

[0217] In the above-described cycles ①, ②, and ③ processes, a memory can be involved to provide data storage support for each agent. For example, in the cyclic process between the application network and the execution agent, the execution data generated by the application network is collected into the memory, and the execution agent adjusts the execution process of the application network based on the stored data.

[0218] The above embodiments introduce the specific process of the network management method proposed in the present application. In the execution process of the above method, two network entities, a memory and a tool library, can be involved. The two network entities are described in detail as follows.

[0219] For the memory:

[0220] (1) The multi-agent system further includes a first memory, and the first memory includes information corresponding to the stored intention request and the first task.

[0221] The foregoing embodiments describe that the intent request is input into the network model, and a corresponding first task can be output. Then the first memory can correspondingly store information of the intent request and information of the first task set, so that when the same intent request is received again, the first task set can be directly obtained according to the corresponding stored content in the memory. Alternatively, the information of the intent request can be complete intent request content, or can be a keyword or abstract content of the intent request. In order to reduce the storage amount of the memory. Similarly, the information of the first task set can be identification information of the first task set, or identification information of a subtask in the first task set, etc.

[0222] In addition, when the intent request is input into the network model, other information such as related context, historical data, or prompts needs to be input into the network model to obtain the output result. These data can come from the zeroth memory.

[0223] (2) The multi-agent system further includes a second memory, and the second memory includes information of the corresponding stored first task and information of the first function module set.

[0224] The foregoing embodiments describe that each subtask in the first task needs to be matched to a function module to obtain the first function module set. Then the information of the first task and the information of the first function module set can be correspondingly stored in the second memory. Including correspondingly storing the first task identification and the first function module set identification, or correspondingly storing the identification of each subtask in the first task and the identification of the corresponding function module in the first function module set, etc. So that when the first task is obtained next time, the first function module set can be directly obtained from the second memory. Improve the efficiency of obtaining.

[0225] (3) The multi-agent system further includes a third memory, and the third memory includes information of the corresponding stored first network topology and resource requirement information of the function module of the first network topology.

[0226] The foregoing embodiments describe that the first network topology can further include resource requirements of each function module in the network topology. Then the information of the first network topology and the resource requirements of each function module in the first network topology can be correspondingly stored, so that when the same network topology is obtained next time, the resource requirements of the function module can be directly determined. The first network topology information can be the function module name included in the network topology, or can be the network topology identification, etc.

[0227] Or the actual memory included in the multi-agent system can be more than the above-mentioned memory quantity, or can be less than the above-mentioned memory quantity. In addition, these memories can be shared memories, or can be independent memories. Specifically, refer to FIG. 5A. FIG. 5A is a schematic diagram of a network entity layout provided by an embodiment of the present application. As shown in FIG. 5A, the first memory, the second memory, the third memory, and even the zeroth memory are shared memories. Therefore, the multi-agent system, i.e., the orchestration agent, the connection agent, the assembly agent, and the execution agent, are connected to one memory, and the above-mentioned data storage support is provided by the memory.

[0228] Alternatively, the first memory, the second memory, the third memory, and the zeroth memory can be independent memories. Specifically, refer to FIG. 5B. FIG. 5B is another schematic diagram of a network entity layout provided by an embodiment of the present application. The zeroth memory (memory 0) can be arranged in the orchestration agent, the first memory (memory 1) can be arranged in the assembly agent, the second memory (memory 2) can be arranged in the execution agent, and the third memory (memory 3) can be arranged in the execution agent. Alternatively, each memory can be independently arranged in other agents, which is not limited in the embodiment of the present application.

[0229] Alternatively, the first memory, the second memory, the third memory, and the zeroth memory can be partially shared memories. Specifically, refer to FIG. 5C. FIG. 5C is another schematic diagram of a network entity layout provided by an embodiment of the present application. The memory 0, the memory 1, and the memory 2 are shared memories, which are represented as the memory 11 in FIG. 5C and are shared by the orchestration agent, the assembly agent, and the execution agent. The third memory in the connection agent is still an independent memory (the memory 12 in the figure).

[0230] For the tool library:

[0231] (1) The multi-agent system includes a first tool library, which is used for the assembly agent to match and obtain a first function module set from the first tool library according to a first task.

[0232] The first tool library can include information of a plurality of function modules. The assembly agent calls a network model, and takes the function module information and the subtask information as inputs, so as to match and obtain the first function module set for the subtask by the network model.

[0233] (2) The multi-agent system further includes a second tool library. The second tool library is used for providing function modules for assembly to the assembly agent.

[0234] The second tool library can include various function module components or functions. The assembly agent encapsulates the function modules from the second tool library for subsequent deployment.

[0235] (3) The multi-agent system further comprises a third tool library, and the third tool library is configured to provide deployment tools for the function modules in the first network topology, including creating a deployment environment, configuring environment parameters, and the like.

[0236] Alternatively, the tool libraries actually included in the multi-agent system can be more than the above tool library quantity, or less than the above tool library quantity. In addition, the tool libraries can be shared tool libraries, or independent tool libraries. For details, refer to the foregoing FIG. 5A. The first tool library, the second tool library, and the third tool library are shared tool libraries, and are connected to the multi-agent system. That is, the orchestration agent, the connection agent, the assembly agent, and the execution agent are connected to one tool library, and all tools are provided by the tool library.

[0237] Alternatively, refer to the foregoing FIG. 5B. When the first tool library, the second tool library, and the third tool library are independent tool libraries, the first tool library (tool library 1) can be located in the orchestration agent. The second tool library (tool library 2) can be located in the assembly agent. The third tool library (tool library 3) can be located in the execution agent. The connection agent can also have an independent tool library 4 for providing function support for the implementation of the connection agent.

[0238] Alternatively, refer to the foregoing FIG. 5C. Part of the tool libraries are shared, and part of the tool libraries are independent. The orchestration agent has an independent tool library 00, and the connection agent has an independent tool library 10. The assembly agent and the execution agent share a tool library 20.

[0239] As can be seen, in the embodiments of the present application, possible layout modes of the memory and the tool library included in the multi-agent system are provided. Independent layout of the memory and / or the tool library can guarantee independent operation of each agent, and does not affect the use of the network entity between the agents, thereby improving the fault tolerance of the multi-agent system. Shared layout of the memory and / or the tool library can reduce the implementation cost and complexity.

[0240] Please refer to FIG. 6, which is a structural schematic diagram of a communication apparatus provided in an embodiment of the present application. The communication apparatus 10 can be used to execute any of the methods in the foregoing embodiments.

[0241] As shown in FIG. 6, the communication apparatus comprises a processing module 1501 and a transceiver module 1502. The processing module 1501 can be one or more processors, and the transceiver module 1502 can be a transceiver or a communication interface. The communication apparatus can be used to implement the functions of the agents in any of the method embodiments. These apparatuses can be hardware devices, software functions running on special hardware, or virtualized functions instantiated on a platform (for example, a cloud platform). Optionally, the communication apparatus can further comprise a storage module 1503 configured to store program codes and data of the communication apparatus.

[0242] In the first example, the communication device can be used as a chip in the agents (orchestration agent, assembly agent, execution agent and connection agent) in FIGS. 3A-5C, and perform the steps performed by the agents in the method embodiments described above. The transceiver module 1502 is configured to support communication between the agents, or between the agents and the external communication devices outside the multi-agent system. The processing module 1501 can be configured to support the actions performed by the agents in the method embodiments described above, except for sending and receiving.

[0243] For example, when the communication device is used to implement the function of the orchestration agent, the transceiver module 1502 is configured to obtain the intent request.

[0244] The processing module 1501 is configured to obtain the first task according to the intent request and the network model, and the first task includes at least one first subtask and at least one second subtask.

[0245] The transceiver module 1502 is further configured to send the information of the first task to the assembly agent.

[0246] When the communication device is used to implement the function of the assembly agent, the transceiver module 1502 is configured to receive the information of the first task.

[0247] The processing module 1501 is configured to obtain the first function module set according to the at least one first subtask and the first tool library, and the first function module set includes at least one function module corresponding to the at least one first subtask; and obtain the first network topology according to the at least one function module in the first function module set.

[0248] The transceiver module 1502 is configured to send the information of the first network topology to the execution agent.

[0249] When the communication device is used to implement the function of the execution agent, the transceiver module 1502 is configured to receive the information of the first network topology.

[0250] The processing module 1501 is configured to obtain the core network corresponding to the first network topology according to the at least one function module in the first network topology.

[0251] The transceiver module 1502 is further configured to send the connection related information to the connection agent, and the connection related information includes the information of the core network and the task topology information related to the at least one second subtask in the first network topology.

[0252] When the communication device is used to implement the function of the connection agent, the transceiver module 1502 is configured to receive the connection related information.

[0253] The processing module 1501 is configured to configure at least one second sub-task to the communication device according to the task topology information, and establish a communication connection between the core network and the communication device to construct an application network, the communication device including at least one of an access network device or a terminal device.

[0254] The method steps that the communication device can further perform can refer to the related embodiments described above, and will not be described here.

[0255] FIG. 7 is a schematic diagram of another communication device 20 according to an embodiment of the present application. The device 20 includes a processor 21 configured to execute computer programs or instructions stored in a memory 22, or read data / signaling stored in the memory 22, to perform the methods in the above method embodiments. Optionally, the processor 21 is one or more.

[0256] Optionally, as shown in FIG. 7, the device 20 further includes a memory 22 configured to store computer programs or instructions and / or data. The memory 22 can be integrated with the processor 21, or can be separately arranged. Optionally, the memory 22 is one or more.

[0257] Optionally, as shown in FIG. 7, the device 20 further includes a transceiver 23 configured to receive and / or send signals. For example, the processor 21 is configured to control the transceiver 23 to receive and / or send signals.

[0258] As an option, the device 20 is configured to implement the operations performed by the orchestration intelligent agent in the above method embodiments.

[0259] As another option, the device 20 is configured to implement the operations performed by the assembly intelligent agent in the above method embodiments.

[0260] As yet another option, the device 20 is configured to implement the operations performed by the execution intelligent agent in the above method embodiments.

[0261] As yet another option, the device 20 is configured to implement the operations performed by the connection intelligent agent in the above method embodiments.

[0262] It should be understood that the processor mentioned in the embodiments of the present application can be one or a combination of a central processing unit (CPU), other general-purpose processors, a digital signal processor (DSP), a microprocessor unit (MPU), a microcontroller unit (MCU), a GPU, a field programmable gate array (FPGA), an artificial intelligence processor (AI processor), or a neural network processor (Neural Processing Unit, NPU); or the processor mentioned in the embodiments of the present application can be an ASIC or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0263] It should also be understood that the memory mentioned in the embodiments of the present application can be a volatile memory and / or a non-volatile memory. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically EPROM (EEPROM), or a flash memory. The volatile memory can be a cache, a random access memory (RAM). For example, the RAM can be used as an external cache. As an example but not limitation, the RAM includes the following various forms: static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct memory bus random access memory (direct rambus RAM, DR RAM).

[0264] It should be noted that when the processor is a general processor, a DSP, an ASIC, a FPGA or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, the memory (storage module) can be integrated in the processor.

[0265] It should also be noted that the memory described herein is intended to include, but not limited to, these and any other suitable type of memory.

[0266] The embodiments of the present application also provide a chip system, which can also be referred to as a processing system, comprising a logic circuit and an input / output interface.

[0267] The logic circuit can be a processing circuit in the chip system. The logic circuit can be coupled to a storage unit to invoke instructions in the storage unit, so that the chip system can implement the methods and functions of the embodiments of the present application. The input / output interface can be an input / output circuit in the chip system, which outputs information processed by the chip system or inputs data or signaling information to be processed by the chip system.

[0268] As a solution, the chip system is configured to implement the operations performed by the orchestration agent, the assembly agent, the execution agent and the connection agent in the above various method embodiments.

[0269] For example, the logic circuit is configured to implement the processing-related operations performed by the orchestration agent, the assembly agent, the execution agent and the connection agent in the above method embodiments; and the input / output interface is configured to implement the sending and / or receiving-related operations performed by the orchestration agent, the assembly agent, the execution agent and the connection agent in the above method embodiments.

[0270] The embodiments of the present application also provide a computer readable storage medium having stored thereon computer instructions for implementing the methods performed by the orchestration agent, the assembly agent, the execution agent and the connection agent in the above various method embodiments.

[0271] For example, when the computer program is executed by a computer, the computer can implement the methods performed by the orchestration agent, the assembly agent, the execution agent and the connection agent in the above various method embodiments.

[0272] The embodiments of the present application also provide a computer program product comprising instructions, which, when executed by a computer, implement the methods performed by the orchestration agent, the assembly agent, the execution agent and the connection agent in the above various method embodiments.

[0273] The embodiment of the present application further provides a communication system, comprising the foregoing orchestration intelligent entity, assembly intelligent entity, execution intelligent entity and connection intelligent entity. Optionally, the communication system further comprises the foregoing orchestration intelligent entity, assembly intelligent entity, execution intelligent entity and connection intelligent entity, access network equipment and terminal equipment.

[0274] The explanations and beneficial effects of the related contents in any of the above-provided devices can refer to the corresponding method embodiments provided above, and will not be repeated here.

[0275] Those skilled in the art can understand that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized in electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solutions. Those skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0276] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described system, device and unit can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0277] In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division, and actual implementation can have another division manner. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or other forms.

[0278] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e. can be located in one place, or can be distributed on multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0279] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit.

[0280] If the functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the parts that contribute to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of 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 methods described in the various embodiments of the present application. The aforementioned storage medium includes various media that can store program codes, such as a U disk, a mobile hard disk, a ROM, a RAM, a magnetic disk or an optical disk, etc.

[0281] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A network management method characterized by comprising: The method is applied to a multi-agent system, and the multi-agent system includes an orchestration agent, an assembly agent, an execution agent, and a connection agent, wherein The orchestration agent obtains an intention request, obtains a first task according to the intention request and a network model, and sends information of the first task to the assembly agent, the first task including at least one first subtask and at least one second subtask, the first subtask corresponding to a core network, and the second subtask corresponding to a communication device; The assembly agent receives the information of the first task from the orchestration agent, obtains a first function module set according to the at least one first subtask and a first tool library, and the first function module set includes at least one function module corresponding to the at least one first subtask; The assembly agent obtains a first network topology according to the at least one function module and the at least one second subtask, and sends information of the first network topology to the execution agent; The execution agent receives the information of the first network topology from the assembly agent, and obtains a core network corresponding to the first network topology according to at least one function module in the first network topology; The execution agent sends connection-related information to the connection agent, and the connection-related information includes information of the core network and task topology information related to the at least one second subtask in the first network topology; The connection agent receives the connection-related information, configures the at least one second subtask on the communication device according to the task topology information, and establishes a communication connection between the core network and the communication device to construct an application network, and the communication device includes at least one of an access network device or a terminal device.

2. The method of claim 1, wherein, Before the orchestration agent obtains the first task, the method further includes: The orchestration agent obtains a second task according to the intention request and the network model, and sends information of the second task to the assembly agent, the second task including at least one different first subtask from the first task; The assembly agent receives the information of the second task from the orchestration agent, and determines that obtaining a second function module set according to the at least one first subtask in the second task and the first tool library fails; The assembly agent sends a first response to the orchestration agent, and the first response indicates that obtaining the second function module set fails; The orchestration agent receives the first response; The first task is obtained according to the intention request and the network model, including: The orchestration agent obtains the first task according to the second task, the first response, and the network model.

3. The method according to claim 1 or 2, characterized in that, Before the assembly agent obtains the first function module set, the method further includes: The assembly agent obtains a third function module set according to the at least one first subtask in the first task and the first tool library, and the third function module set includes at least one function module corresponding to the at least one first subtask; The assembling agent obtains a second network topology according to at least one function module in the third function module set and the at least one second subtask, and sends information of the second network topology to the executing agent; The executing agent receives the information of the second network topology from the assembling agent, and determines that a core network corresponding to the second network topology fails to be obtained according to at least one function module in the second network topology; The executing agent sends a second response to the assembling agent, and the second response indicates that the core network corresponding to the second network topology fails to be obtained; The assembling agent receives the second response from the executing agent; The first function module set is obtained according to the at least one first subtask and a first tool library, including: The assembling agent obtains a first function module set according to the second response, the at least one first subtask and the first tool library, wherein the third function module set and the first function module set include at least one different function module.

4. The method according to any one of claims 1 to 3, characterized in that, The executing agent obtains a core network corresponding to the first network topology according to at least one function module in the first network topology, including: The executing agent deploys at least one function module in the first network topology to at least one core network element to obtain a core network composed of the at least one core network element.

5. The method of claim 4, wherein, The first network topology further includes resource requirements of each function module in the at least one function module, and the deploying of the at least one function module in the first network topology to at least one core network element includes: Deploying the at least one function module in the first network topology to at least one core network element according to the resource requirements of each function module.

6. The method of claim 3, wherein, The executing agent determines that the core network corresponding to the second network topology fails to be obtained according to at least one function module in the second network topology, including that the executing agent determines that at least one function module in the second network topology includes a function module that fails to be deployed, and the function module that fails to be deployed includes at least one of the following: The core network does not meet the resource requirements of the function module; or The resources of the core network do not meet the quality of service (QoS) requirements of the function module.

7. The method according to any one of claims 1 to 6, characterized in that, The method further includes: The executing agent obtains network state information of the application network; When the network state information indicates that the application network performs abnormally, the executing agent adjusts the application network, including at least one of the following: At least one of the following: adjusting the number of the at least one function module and adjusting the position of the at least one function module; At least one of the following: adjusting the number of the at least one second subtask and adjusting the position of the at least one second subtask.

8. The method according to any one of claims 1 to 7, characterized in that, The multi-agent system further includes a first storage, and the first storage includes information corresponding to the intention request and information of the first task stored therein; or The multi-agent system further includes a second storage, and the second storage includes information corresponding to the first task and information of the first function module set stored therein; or ​ The multi-agent system further comprises a third memory, and the third memory comprises corresponding stored information of the first network topology and resource requirement information of functional modules of the first network topology.

9. The method of claim 8, wherein, The first memory, the second memory and the third memory are shared memories; or Any two of the first memory, the second memory and the third memory are shared memories; or None of the first memory, the second memory and the third memory are shared memories. The multi-agent system further comprises a second tool library, and the second tool library is used to provide the functional modules for assembly for the assembly agent; and / or 10. The method according to any one of claims 1 to 9, characterized in that, The multi-agent system further comprises a third tool library, and the third tool library is used to provide tools for deployment for the functional modules in the first network topology. The first tool library, the second tool library and the third tool library are shared tool libraries; or 11. The method of claim 10, wherein, Any two of the first tool library, the second tool library and the third tool library are shared tool libraries; or None of the first tool library, the second tool library and the third tool library are shared tool libraries. The multi-agent system comprises an orchestration agent, an assembly agent, an execution agent and a connection agent, wherein 12. A multi-agent system, characterized in that, The orchestration agent is configured to obtain an intent request, obtain a first task according to the intent request and a network model, and send information of the first task to the assembly agent, the first task comprising at least one first subtask and at least one second subtask, the first subtask corresponding to a core network, and the second subtask corresponding to a communication device; The assembly agent is configured to receive the information of the first task from the orchestration agent, obtain a first functional module set according to the at least one first subtask and a first tool library, and the first functional module set comprising at least one functional module corresponding to the at least one first subtask; The assembly agent is further configured to obtain a first network topology according to the at least one functional module and the at least one second subtask, and send information of the first network topology to the execution agent; The execution agent is configured to receive the information of the first network topology from the assembly agent, obtain a core network corresponding to the first network topology according to at least one functional module in the first network topology; The execution agent is further configured to send connection related information to the connection agent, the connection related information comprising information of the core network and task topology information of the first network topology related to the at least one second subtask; The connection agent is configured to receive the connection related information, configure the at least one second subtask to the communication device according to the task topology information, and establish a communication connection between the core network and the communication device to construct an application network, the communication device comprising at least one of an access network device or a terminal device.

13. The system of claim 12, wherein ​ The orchestration agent is further configured to receive information of the second task from the orchestration agent, and determine that obtaining a second function module set according to at least one first subtask in the second task and a first tool library fails. The assembly agent is further configured to receive information of the second task from the orchestration agent, and determine that obtaining a second function module set according to at least one first subtask in the second task and a first tool library fails. The assembly agent is further configured to send a first response to the orchestration agent, the first response indicating that obtaining the second function module set fails. The orchestration agent is further configured to receive the first response. The first task is obtained according to the intent request and a network model, and the obtaining includes: The orchestration agent is further configured to obtain the first task according to the second task, the first response and the network model.

14. The system of claim 12 or 13, wherein, The assembly agent is further configured to obtain a third function module set according to at least one first subtask in the first task and the first tool library, the third function module set including at least one function module corresponding to the at least one first subtask. The assembly agent is further configured to obtain a second network topology according to at least one function module in the third function module set and the at least one second subtask, and send information of the second network topology to an execution agent. The execution agent is further configured to receive information of the second network topology from the assembly agent, and determine that obtaining a core network corresponding to the second network topology according to at least one function module in the second network topology fails. The execution agent is further configured to send a second response to the assembly agent, the second response indicating that obtaining the core network corresponding to the second network topology fails. The assembly agent receives the second response from the execution agent. The first function module set is obtained according to the at least one first subtask and the first tool library, and the obtaining includes: The assembly agent is further configured to obtain a first function module set according to the second response, the at least one first subtask and the first tool library, wherein the third function module set and the first function module set include at least one different function module.

15. The system according to any of claims 12-14, characterized in that, The execution agent obtains the core network corresponding to the first network topology according to at least one function module in the first network topology, and the obtaining includes: The execution agent deploys at least one function module in the first network topology to at least one core network element to obtain a core network composed of the at least one core network element.

16. The system of claim 15, wherein, The first network topology further includes resource requirements of each function module in the at least one function module, and the deploying includes: deploying the at least one function module in the first network topology to at least one core network element according to the resource requirements of each function module.

17. The system of claim 14, wherein, The execution agent determines that the at least one functional module in the second network topology includes a functional module with a deployment failure, and the functional module deployment failure includes at least one of the following: The core network does not meet the resource requirement of the functional module; or The resource of the core network does not meet the quality of service (QoS) requirement of the functional module.

18. The system of any one of claims 12-17, wherein: The execution agent is further configured to obtain network state information of the application network. When the network state information indicates that the application network performs abnormally, the execution agent is further configured to adjust the application network, including at least one of the following: At least one of the following: adjusting the number of the at least one functional module and adjusting the position of the at least one functional module; and Indicating the connection agent to perform at least one of the following: adjusting the number of the at least one second subtask and adjusting the position of the at least one second subtask.

19. The system of any of claims 12-18, wherein, The multi-agent system further comprises a first storage, wherein the first storage comprises information corresponding to the information of the intent request and the information of the first task stored therein; or The multi-agent system further comprises a second storage, wherein the second storage comprises information corresponding to the information of the first task and the information of the first functional module set stored therein; or The multi-agent system further comprises a third storage, wherein the third storage comprises information corresponding to the information of the first network topology and the resource requirement information of the functional module of the first network topology stored therein. The first storage, the second storage, and the third storage are shared storage; or 20. The system of claim 19, wherein, Any two of the first storage, the second storage, and the third storage are shared storage; or None of the first storage, the second storage, and the third storage is shared storage. The multi-agent system further comprises a second tool library, wherein the second tool library is configured to provide the functional module for assembly for the assembly agent; and / or The multi-agent system further comprises a third tool library, wherein the third tool library is configured to provide a tool for deployment for the functional module in the first network topology.

21. The system of any of claims 12-20, wherein, The first tool library, the second tool library, and the third tool library are shared tool libraries; or Any two of the first tool library, the second tool library, and the third tool library are shared tool libraries; or 22. The system of claim 21, wherein, None of the first tool library, the second tool library, and the third tool library is shared tool library. The apparatus comprises a module or unit for implementing the method of any one of claims 1-11.

24. A communication apparatus, comprising:

23. A communications device, characterized by at least one processor coupled with a memory; wherein the at least one processor is configured to execute computer programs or instructions stored in the memory, so that the method performed by any agent in the method of any one of claims 1-11 is implemented. ​ ​ 25. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program which, when executed, causes the method performed by any of the agents in claims 1 to 11 to be implemented.

26. A computer program, characterized in that, The computer program, when executed, causes the method performed by any of the agents in claims 1 to 11 to be implemented.

27. A computer program product, characterised in that, The computer program product comprises one or more computer programs which, when executed on a computer, cause the computer to perform the method of any of claims 1 to 11.

28. A chip, characterized by A processor configured to perform the method of any of claims 1 to 11.

Images