Communication method and apparatus for agent network, and system

Abstract

Embodiments of the present application relate to the technical field of artificial intelligence, and provide a communication method and apparatus for an agent network, and a system. The method comprises: an agent instance corresponding to a first terminal receives indication information from the first terminal. The indication information is used for indicating an intention of a user. On the basis of the indication information, the agent instance corresponding to the first terminal determines a target agent instance. The capability of the target agent instance comprises the capability of a terminal corresponding to the target agent instance and a capability assigned by a network to the terminal. The agent instance corresponding to the first terminal triggers the target agent instance to execute a task on the basis of the capability of the target agent instance so as to achieve an intention. Since a target agent instance stores the capability of a terminal and a capability assigned by a network to the terminal over a long period, and the capability assigned by the network to the terminal can perform combined planning on the capability of the terminal, an agent instance corresponding to a first terminal can execute a complex intelligent operation by using a target agent network at any time.

Description

A communication method, device and system of an agent network

[0001] The present application claims priority to the Chinese patent application No. 202410956543.4, filed on July 16, 2024, and entitled "A communication method, device and system of an agent network", the content of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0002] The present application relates to the field of artificial intelligence, and in particular, to a communication method, device and system of an agent network. BACKGROUND

[0003] An "agent" refers to a computing entity that resides in a certain environment, can continuously and autonomously function, and has characteristics such as residency, reactivity, sociality, and initiative. In an agent network composed of multiple different agents, since the agent network has the characteristics of being compatible with multiple communication standards and interfaces, the carriers (such as terminals) of different agents have end-to-end (E2E) interoperability. An agent can access the agent network and establish a session to establish communication with another agent.

[0004] In the prior art, multiple agents in an agent network cannot cooperate together to complete the user's intention, and cannot effectively call the capabilities of the agents given by the network to realize complex intelligent operations among multiple agents. SUMMARY

[0005] The present application provides a communication method, device and system of an agent network to realize multi-agent cooperation to complete the user's intention.

[0006] The technical solution is as follows:

[0007] In a first aspect, the embodiments of the present application provide a communication method of an agent network, the method comprising: an agent instance corresponding to a first terminal receiving indication information from the first terminal. The indication information is used to indicate the intention of a user, and the user is associated with the first terminal. The agent instance corresponding to the first terminal determines a target agent instance according to the indication information. The capability of the target agent instance includes the capability of a terminal corresponding to the target agent instance and the capability given to the terminal by the network. The agent instance corresponding to the first terminal triggers the target agent instance to perform a task according to the capability of the target agent instance to achieve the intention.

[0008] The application receives the indication information of the first terminal through the agent instance corresponding to the first terminal, and indicates the intention of the user. Then, the target agent instance required to achieve the intention is determined according to the intention. Since the target agent instance includes the capability of the terminal corresponding to the agent instance and the capability of the network for the terminal, the agent instance corresponding to the first terminal can trigger the target agent instance to execute the task according to the target agent instance. In this way, since the capability of the terminal and the capability of the network for the terminal are long-term stored in the target agent instance, the agent instance corresponding to the first terminal can execute complex intelligent operations at any time by using the target agent network.

[0009] In a possible implementation, the agent instance corresponding to the first terminal determines the target agent instance according to the indication information, including: the agent instance corresponding to the first terminal determines the task according to the intention of the user. The agent instance corresponding to the first terminal determines the target agent instance including the agent instance with the capability for executing the task.

[0010] In a possible implementation, the task includes a first subtask, and the agent instance corresponding to the first terminal determines the target agent instance including the agent instance with the capability for executing the task, including: the agent instance corresponding to the first terminal determines the capability for executing the first subtask. The agent instance corresponding to the first terminal determines the target agent instance including the first agent instance. The first agent instance has the capability for executing the first subtask.

[0011] In a possible implementation, the agent instance corresponding to the first terminal determines the target agent instance including the first agent instance, including: the agent instance corresponding to the first terminal sends a request message to the agent registration function network element. The request message is used to indicate the capability for executing the first subtask. The agent instance corresponding to the first terminal receives the information of the first agent instance from the agent registration function network element.

[0012] In a possible implementation, the agent instance corresponding to the first terminal determines the target agent instance including the first agent instance, including: the agent instance corresponding to the first terminal determines the first agent instance from at least one agent instance associated with the agent instance corresponding to the first terminal.

[0013] In a possible implementation, the at least one agent instance associated with the agent instance corresponding to the first terminal includes: the terminal corresponding to the at least one agent instance belongs to the user with the first terminal.

[0014] In a possible implementation, the indication information includes an identifier of the second terminal, and the agent instance corresponding to the first terminal determines the target agent instance according to the indication information, including: the agent instance corresponding to the first terminal determines the target agent instance including the agent instance corresponding to the second terminal according to the identifier of the second terminal.

[0015] In a possible implementation, the capability given to the terminal by the network is used to plan and combine the capability of the terminal.

[0016] In a possible implementation, the agent instance corresponding to the first terminal triggers the target agent instance to perform a task according to the capability of the target agent instance to achieve the intent, including: the agent instance corresponding to the first terminal sends information describing the first subtask to the target agent instance.

[0017] In a possible implementation, the method provided by the embodiment of the application further includes: the agent instance corresponding to the first terminal receives information describing the first subtask. The agent instance corresponding to the first terminal invokes the capability of the target agent instance to perform the first subtask in response to the information describing the first subtask.

[0018] In a possible implementation, the agent instance corresponding to the first terminal invokes the capability of the target agent instance to perform the first subtask, including: the agent instance corresponding to the first terminal plans and combines the capability of the terminal according to the capability given to the terminal by the network to perform the first subtask.

[0019] In a second aspect, the embodiment of the application provides a communication method of an agent network, including: an artificial intelligence agent function network element creates an agent instance corresponding to a first terminal. The artificial intelligence agent function network element instructs the agent instance to send a registration request message to an agent registration function network element. The registration request message includes the capability of the agent instance, and the capability of the agent instance includes the capability of the first terminal and the capability given to the first terminal by the network.

[0020] In a possible implementation, before creating the agent instance corresponding to the first terminal, the method provided by the embodiment of the application further includes: the artificial intelligence agent function network element receives information indicating the capability of the first terminal from the first terminal. The artificial intelligence agent function network element creates the agent instance corresponding to the first terminal, including: the artificial intelligence agent function network element creates the agent instance corresponding to the first terminal according to the capability of the first terminal.

[0021] In a possible implementation, the artificial intelligence agent function network element creates the agent instance corresponding to the first terminal according to the capability of the first terminal, including: the artificial intelligence agent function network element determines an agent template according to the capability of the first terminal and the subscription information of the first terminal. The artificial intelligence agent function network element creates the agent instance corresponding to the first terminal according to the agent template.

[0022] In a possible implementation, the method provided by the embodiment of the application further includes: the artificial intelligence agent function network element receives information indicating the capability of the second terminal from the second terminal. The artificial intelligence agent function network element creates the agent instance corresponding to the user in the case that the first terminal and the second terminal belong to the same user. The capability of the agent instance corresponding to the user includes the capability of the agent instance corresponding to the first terminal and the capability of the agent instance corresponding to the second terminal.

[0023] In a possible implementation, the method provided by the embodiment of the application further includes: the artificial intelligence agent function network element instructs the agent instance corresponding to the user to send a second request message to the agent registration function network element. The second request message is used to indicate the capability of the agent instance corresponding to the user.

[0024] In a possible implementation, the method provided by the embodiment of the application further includes: the artificial intelligence agent function network element instructs the agent instance corresponding to the user to send a second request message to the agent registration function network element. The second request message is used to indicate the capability of the agent instance corresponding to the user.

[0025] In a possible implementation, the method provided by the embodiment of the application includes: the agent registration function network element receives a registration request message from the agent instance corresponding to the first terminal. The registration request message includes the capability of the agent instance.

[0026] In a possible implementation, the method provided by the embodiment of the application further includes: the agent registration function network element receives a second request message from the agent instance corresponding to the user. The second request message is used to indicate the capability of the agent instance corresponding to the user. The capability of the agent instance corresponding to the user includes the capability of the agent instance corresponding to the first terminal and the capability of the agent instance corresponding to the second terminal.

[0027] Fourthly, embodiments of this application provide a communication device that can implement the methods in the first aspect or any possible implementation of the first aspect, and therefore can also achieve the beneficial effects of the first aspect or any possible implementation of the first aspect. This communication device can be an intelligent agent instance, or an apparatus that supports the intelligent agent instance in implementing the methods in the first aspect or any possible implementation of the first aspect, such as a chip applied in the intelligent agent instance. This device can implement the above methods through software, hardware, or by hardware executing corresponding software.

[0028] Fifthly, embodiments of this application provide a communication device that can implement the methods in the second aspect or any possible implementation of the second aspect, and therefore can also achieve the beneficial effects of the second aspect or any possible implementation of the second aspect. This communication device can be a network device, or an apparatus that supports a network device in implementing the methods in any possible implementation of the second aspect or the first aspect, such as a chip applied in an artificial intelligence agent function network element. This device can implement the above methods through software, hardware, or hardware executing corresponding software.

[0029] Sixthly, embodiments of this application provide a communication device that can implement the methods in the third aspect or any possible implementation of the third aspect, and therefore can also achieve the beneficial effects of the third aspect or any possible implementation of the third aspect. This communication device can be a network device, or an apparatus that supports a network device in implementing the methods in the third aspect or any possible implementation of the third aspect, such as a chip applied in a network element for an intelligent agent registration function. This device can implement the above methods through software, hardware, or by hardware executing corresponding software.

[0030] In a seventh aspect, embodiments of this application provide a computer-readable storage medium storing a computer program or instructions that, when executed on a computer, cause the computer to perform the methods described in any of the possible implementations of the first aspect.

[0031] Eighthly, embodiments of this application provide a computer-readable storage medium storing a computer program or instructions that, when executed on a computer, cause the computer to perform the methods described in any of the possible implementations of the second aspect.

[0032] Ninthly, embodiments of this application provide a computer-readable storage medium storing a computer program or instructions that, when executed on a computer, cause the computer to perform the methods described in any of the possible implementations of the third aspect to the third aspect.

[0033] In a tenth aspect, embodiments of this application provide a computer program product including instructions that, when executed on a computer, cause the computer to perform a data transmission method described in the first aspect or various possible implementations of the first aspect.

[0034] Eleventhly, embodiments of this application provide a computer program product including instructions that, when executed on a computer, cause the computer to perform a data transmission method described in the second aspect or various possible implementations of the second aspect.

[0035] In a twelfth aspect, embodiments of this application provide a computer program product including instructions that, when executed on a computer, cause the computer to perform a data transmission method described in the third aspect or various possible implementations of the third aspect.

[0036] In a thirteenth aspect, embodiments of this application provide a communication device comprising: at least one processor. The at least one processor is coupled to a memory, and when the communication device is in operation, the processor executes computer execution instructions or programs stored in the memory to cause the communication device to perform a method as described in the first aspect or any of the various possible designs of the first aspect.

[0037] In a fourteenth aspect, embodiments of this application provide a communication device comprising: at least one processor. The at least one processor is coupled to a memory, and when the communication device is in operation, the processor executes computer execution instructions or programs stored in the memory to cause the communication device to perform a method as described in the second aspect above, or any of the various possible designs of the second aspect.

[0038] In a fifteenth aspect, embodiments of this application provide a communication device comprising: at least one processor. The at least one processor is coupled to a memory, and when the communication device is in operation, the processor executes computer execution instructions or programs stored in the memory to cause the communication device to perform a method as described in the third aspect above, or any of the various possible designs of the third aspect.

[0039] It should be understood that the memory described in any of aspects thirteen to fifteen can also be replaced by a storage medium, and the embodiments of this application do not limit this.

[0040] In one possible implementation, the memory described in any one of aspects ten to eleven can be a memory inside the communication device. Of course, the memory can also be located outside the communication device, but at least one processor can still execute computer execution instructions or programs stored in the memory.

[0041] In a sixteenth aspect, embodiments of this application provide a communication device comprising one or more modules for implementing the methods of any one of the first, second, and third aspects described above. The one or more modules may correspond to the various steps in the methods of any one of the first, second, and third aspects described above.

[0042] In a seventeenth aspect, embodiments of this application provide a chip system including a processor. The processor reads and executes a computer program stored in a memory to perform the methods of the first aspect and any possible implementation thereof. Optionally, the chip system may be a single chip or a chip module composed of multiple chips. Optionally, the chip system further includes a memory, which is connected to the processor via circuitry or wiring. Further optionally, the chip system includes a communication interface. The communication interface is used to communicate with other modules outside the chip.

[0043] In an eighteenth aspect, embodiments of this application provide a chip system including a processor. The processor reads and executes a computer program stored in a memory to perform the methods in the second aspect and any possible implementation thereof. Optionally, the chip system may be a single chip or a chip module composed of multiple chips. Optionally, the chip system further includes a memory, which is connected to the processor via circuitry or wiring. Further optionally, the chip system includes a communication interface. The communication interface is used to communicate with other modules outside the chip.

[0044] In a nineteenth aspect, embodiments of this application provide a chip system including a processor. The processor reads and executes a computer program stored in a memory to perform the methods of the third aspect and any possible implementation thereof. Optionally, the chip system may be a single chip or a chip module composed of multiple chips. Optionally, the chip system further includes a memory, which is connected to the processor via circuitry or wiring. Further optionally, the chip system includes a communication interface. The communication interface is used to communicate with other modules outside the chip.

[0045] In a twentieth aspect, embodiments of this application provide a communication system, comprising: an artificial intelligence agent function network element and an intelligent agent registration function network element. The artificial intelligence agent function network element includes an intelligent agent instance corresponding to a first terminal. The intelligent agent instance corresponding to the first terminal is used to execute the methods in the first aspect and any possible implementation thereof. The artificial intelligence agent function network element is used to execute the methods in the second aspect and any possible implementation thereof. The intelligent agent registration function network element is used to execute the methods in the third aspect and any possible implementation thereof.

[0046] Any of the devices, computer storage media, computer program products, chips, or communication systems provided above are used to execute the corresponding methods provided above. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects of the corresponding solutions in the corresponding methods provided above, and will not be repeated here. Attached Figure Description

[0047] Figure 1 is a conceptual diagram of an agent network provided in an embodiment of this application;

[0048] Figure 2 is a schematic diagram of a networked drone and robot dog collaborating through an intelligent agent network to find lost items, according to an embodiment of this application.

[0049] Figure 3 is an example diagram of finding lost items provided in an embodiment of this application;

[0050] Figure 4 is a schematic diagram of a system architecture provided in an embodiment of this application;

[0051] Figure 5 is a schematic diagram of a communication method for an intelligent agent network provided in an embodiment of this application;

[0052] Figure 6 is a schematic diagram of another communication method for an intelligent agent network provided in an embodiment of this application;

[0053] Figure 7 is a schematic diagram of a terminal creating a corresponding intelligent agent instance according to an embodiment of this application;

[0054] Figure 8 is a schematic diagram of a process provided by an embodiment of this application, in which a third party provides an agent template to an agent network;

[0055] Figure 9 is a flowchart illustrating an OAM configuration agent template provided in an embodiment of this application;

[0056] Figure 10 is a flowchart illustrating a method provided in this application where a smart agent instance corresponding to a first terminal performs a task to achieve an intent initiated by the first terminal.

[0057] Figure 11 is a schematic diagram of a specific implementation method for a terminal to create a corresponding intelligent agent instance according to an embodiment of this application;

[0058] Figure 12 is a schematic diagram of a specific implementation method of a terminal registering an intelligent agent instance according to an embodiment of this application;

[0059] Figure 13 is a schematic diagram of another specific implementation method for creating a corresponding intelligent agent instance by a terminal according to an embodiment of this application;

[0060] Figure 14 is a schematic diagram of another specific implementation method of terminal registering intelligent agent instances provided in the embodiments of this application;

[0061] Figure 15 is a schematic diagram of a specific implementation method of a first terminal corresponding to an intelligent agent instance executing a task through a target intelligent agent instance according to the terminal intent provided in an embodiment of this application;

[0062] Figure 16 is a schematic diagram of a terminal belonging to the same user creating a corresponding intelligent agent instance according to an embodiment of this application;

[0063] Figure 17 is a flowchart illustrating a method for another intelligent agent instance corresponding to a first terminal to perform a task to achieve an intent initiated by the first terminal, according to an embodiment of this application.

[0064] Figure 18 is a schematic diagram of a user-corresponding intelligent agent instance provided in an embodiment of this application;

[0065] Figure 19 is a schematic diagram of a specific implementation method for a subsequent terminal to create a corresponding intelligent agent instance according to an embodiment of this application;

[0066] Figure 20 is a schematic diagram of a specific implementation method for a subsequent terminal to register an intelligent agent instance according to an embodiment of this application;

[0067] Figure 21 is a schematic diagram of another specific implementation method for creating an intelligent agent instance by a subsequent terminal according to an embodiment of this application;

[0068] Figure 22 is a schematic diagram of another specific implementation method of subsequent terminal registering and updating intelligent agent instance provided in the embodiments of this application;

[0069] Figure 23 is a schematic diagram of another specific implementation method of the intelligent agent instance corresponding to the first terminal executing a task through the target intelligent agent instance according to the terminal intent provided in the embodiment of this application;

[0070] Figure 24 is a schematic diagram of a communication device provided in an embodiment of this application;

[0071] Figure 25 is a schematic diagram of the hardware structure of a communication device provided in an embodiment of this application;

[0072] Figure 26 is a schematic diagram of a chip structure provided in an embodiment of this application. Detailed Implementation

[0073] The technical solutions of the embodiments of this application will be described below with reference to the accompanying drawings. In the description of the embodiments of this application, unless otherwise stated, " / " means "or," for example, A / B can mean A or B; "and / or" in this text is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Furthermore, in the description of the embodiments of this application, "multiple" refers to two or more than two.

[0074] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this embodiment, unless otherwise stated, "a plurality of" means two or more.

[0075] Furthermore, the terms “comprising” and “having”, and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus.

[0076] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a specific manner to facilitate understanding.

[0077] It should be understood that in this application, "at least one (item)" means one or more. "More than one" means two or more. "At least two (items)" means two or three or more. "And / or" is used to describe the relationship between related objects, indicating that there can be three relationships. For example, "A and / or B" can mean: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural.

[0078] The character " / " generally indicates that the preceding and following objects are in an "or" relationship. "At least one of the following" or similar expressions refer to any combination of these items, including any single or multiple items. For example, "at least one of a, b, or c" can be expressed as: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0079] Both "...when" and "if" indicate that a corresponding action will be taken under certain objective circumstances. They are not time limits, nor do they require a judgment action to be taken when the action is taken, nor do they imply any other limitations.

[0080] The steps involved in the communication method of an intelligent agent network provided in this application embodiment are merely examples. Not all steps are mandatory, nor are all contents of each piece of information or message required. They can be added or removed as needed during use.

[0081] In this application, the same step or a step or message with the same function can be referenced and learned from each other in different embodiments.

[0082] Before introducing the embodiments of this application, the relevant terms involved in the embodiments of this application are first defined as follows:

[0083] 1. Intelligent Agent: A computational entity that resides in a specific environment and can function autonomously and continuously. It possesses characteristics such as residency, responsiveness, sociality, and initiative. An intelligent agent can be any entity with a certain level of intelligence, such as software programs, robots, or sensors.

[0084] 2. Large Language Models (LLM): These are deep learning models trained on massive amounts of text data. They can not only generate natural language text, but also deeply understand the meaning of text and handle various natural language tasks, such as text summarization, question answering, and translation.

[0085] Referring to Figure 1, which is a conceptual diagram of an agent network provided in an embodiment of this application, Figure 1(a) is a conceptual diagram of the agent network. The agent network consists of multiple agents combined with large language models (LLMs).

[0086] Agents can perform functions such as planning, actions, tools, memory, and reflection. Each agent in a multi-agent system can independently perceive its environment, make decisions, and execute actions. Agents can communicate and collaborate to provide readily available, on-demand intelligent connectivity and information exchange services.

[0087] LLM can process language, encoding, images, videos, robots, and more. Based on this, intelligent agent networks can achieve intelligent connections with humans, machines, digital avatars, and embodied intelligence to complete information exchange.

[0088] As shown in Figure 1(b), humans, machines, digital assistants, and embodied intelligence can autonomously access the agent network based on LLM. Once an agent establishes a session, it can communicate with other agents as needed, and its network behavior is secure and controllable. Humans (also known as natural persons) can interact with the agent network through terminals using natural language (such as text and voice), video, images, and touch; machines can interact with the agent network using machine commands and sensor data; digital assistants can interact with the agent network using intent expressions and semantic features; and embodied intelligence can interact with the agent network using model parameters and intelligent tasks. Centered on the agent network, autonomous and efficient collaboration among agents of different categories, manufacturers, models, and intelligence levels can be achieved to jointly complete tasks. This enables efficient information transmission and accurate understanding of commands between agents to ensure efficient task closure, and provides necessary perception, intelligence, data empowerment, and support services for various agents.

[0089] For example, Figure 2 illustrates a networked drone and robot dog collaborating through an intelligent agent network to locate lost items. The specific method is implemented through specific network elements or devices within the intelligent agent network; here, a specific network element is used as an example. The specific process includes:

[0090] Step 1: The user (e.g., a person) sends a request message to a specific network element through a communication device (e.g., a terminal), and the specific network element receives the request message from the communication device.

[0091] Request messages are used to indicate the user's intent. For example, a request message might be used to request the search for a lost item (such as a wallet) in a predefined area (such as a park).

[0092] For example, a request message can be either voice or an image.

[0093] Step 2: A specific network element uses LLM to identify the user's intent in the request message and determines the task to be performed based on the intent.

[0094] For example, a specific network element can identify the intent in the request message (to find a lost item in a preset area) and determine that a lost item needs to be found in a preset area.

[0095] Step 3: A specific network element determines at least one subtask based on the task to be performed.

[0096] For example, finding lost items in a preset area can be broken down into subtask one (locating the lost items in the preset area) and subtask two (picking up the lost items based on their location).

[0097] Step 4: A specific network element determines the capability required to execute a subtask based on at least one subtask.

[0098] For example, based on sub-task one (locating lost items in a preset area), it is determined that the ability to fly and take pictures is required; based on sub-task two (picking up lost items based on location), the ability to move and grasp items is required.

[0099] Step 5: The specific network element determines the machine with the capability required to execute the subtask.

[0100] For example, based on flight capabilities and photography capabilities, a networked drone can be identified as the machine performing sub-task one; based on mobility and the ability to pick up objects, a robot dog can be identified as the machine performing sub-task two.

[0101] Step 6: A specific network element sends instruction information to the machine designated to perform the task, based on at least one sub-task. Correspondingly, the machine performing the task receives the instruction information from the specific network element. The instruction information is used to instruct the machine to perform an action.

[0102] For example, referring to Figure 3, a specific network element instructs the connected drone to patrol the park along a set path according to subtask one, and to identify and locate lost items; after locating the lost items, the specific network element instructs the robot dog to reach the located location and pick up the lost items according to subtask two.

[0103] Step 7: The specific network element sends a result report to the user based on the task execution status.

[0104] For example, the results report may include: the location of the lost item, the time taken to complete the task, the battery level of the connected drone, and the battery level of the robot dog.

[0105] In existing technologies, while agent networks can enable different terminals to jointly execute tasks based on user intentions, current agent networks cannot store the capabilities of these terminals. Therefore, whenever a terminal needs to perform a task, its capabilities must be retrieved in real time. For more complex tasks, more terminals may be required, and the capabilities granted to the terminals by the network will also be utilized. However, current agent networks cannot effectively utilize the capabilities of various terminals or the capabilities granted to them by the network to achieve complex operations between multiple agents.

[0106] Based on this, embodiments of this application provide a communication method for an intelligent agent network. This method introduces an artificial intelligence agent function, enabling it to create intelligent agent instances for terminals based on intelligent agent templates and according to received terminal capabilities, network capabilities, and subscription information. The capabilities of the intelligent agent instance include terminal capabilities and network empowerment, achieving the effect of encapsulating terminal capabilities and network empowerment and providing them to the outside world in a unified manner.

[0107] The technical solution of this application can be applied to various communication systems, such as: Long Time Evolution (LTE) systems, LTE Frequency Division Duplex (FDD) systems, LTE Time Division Duplex (TDD) systems, Universal Mobile Telecommunication System (UMTS), Worldwide Interoperability for Microwave Access (WiMAX) communication systems, Public Land Mobile Network (PLMN) systems, Device to Device (D2D) network systems or Machine to Machine (M2M) network systems, the 5th Generation (5G) system, and future network communication technologies, etc.

[0108] The network architecture and business scenarios described in the embodiments of this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided in the embodiments of this application. As those skilled in the art will know, with the evolution of network architecture and the emergence of new business scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0109] As shown in Figure 4, Figure 4 is a schematic diagram of a system architecture provided in an embodiment of this application. The system includes: a core network (CN), an access network (AN), and a terminal.

[0110] The core network is primarily responsible for handling the network's main service processes and data transmission, including: agent network, user plane, and control plane.

[0111] The intelligent agent network includes: intelligent session function (ISF) network element, artificial intelligence agent function (AIAF) network element, agent registration function (ARF) network element, and agent template repository (ATR).

[0112] ISF network elements are used to support terminals in creating intelligent sessions, selecting the AIAF function, and requesting the creation of the Agent instance corresponding to the terminal.

[0113] AIAF network elements are used for the creation and management of intelligent agent instances.

[0114] ARF network elements are used to store and register agent instances.

[0115] ATR is used to store agent templates. ART can be integrated into AIAF network elements.

[0116] The user plane includes: user plane function (UPF) network elements and at least one AIAF network element.

[0117] UPF network elements, as interfaces with the data network, perform functions such as user plane data forwarding, session / flow-based billing and statistics, and bandwidth limiting.

[0118] Control plane: Primarily responsible for transmitting control signaling to control the establishment, maintenance, and release of call flows. Specifically, the control plane carries signaling or control messages.

[0119] The access network is used to connect user terminal equipment to the core network. It is mainly responsible for connecting users' voice, data and other communication services to a wider network to realize the transmission of communication services.

[0120] The access network can be a cellular system related to the 3rd Generation Partnership Project (3GPP), such as 4G, 5G mobile communication systems, NTN (non-terrestrial network) systems, or future-oriented evolution systems. The access network can also be an open radio access network (O-RAN or ORAN), a cloud radio access network (CRAN), or a wireless fidelity (WiFi) system, or a communication system that integrates two or more of the above systems.

[0121] Access networks, sometimes also referred to as access network equipment, radio access network (RAN) entities, or access nodes, constitute a part of a communication system and are used to help terminals achieve wireless access.

[0122] Radio access network equipment: A device deployed in a radio access network to provide wireless communication functions for terminals, such as an evolved node B (eNB) in a long term evolution (LTE) system. eNBs can include various forms of macro base stations, micro base stations (also called small cells), relay stations, access points, wearable devices, and vehicle-mounted equipment. An eNB can also be a transmission and reception point (TRP).

[0123] Wireless access network equipment: A device deployed in a wireless access network that meets 5G standards and provides wireless communication functions for terminals, such as a next-generation base station (g nodeB, gNB). gNBs can include various forms of macro base stations, micro base stations (also known as small stations), relay stations, access points, wearable devices, and vehicle-mounted equipment. gNBs can also be transmission and reception points (TRPs) or transmission measurement functions (TMFs). gNBs can include central units (CUs) and distributed units (DUs) integrated on the gNB.

[0124] In addition, wireless access network equipment can also be a radio network controller (RNC), a radio controller in a cloud radio access network (CRAN) system, a base station controller (BSC), a home base station (e.g., home evolved nodeB, or home node B, HNB), a baseband unit (BBU), a transmitting and receiving point (TRP), a transmitting point (TP), a mobile switching center, a base transceiver station (BTS) in a global system for mobile communication (GSM) or code division multiple access (CDMA) network, a node base station (NB) in a wideband code division multiple access (WCDMA) network, an evolved NB (eNB or eNodeB) in LTE, a base station device in a future network, or an access network device in a future evolved PLMN network, or a wearable device or vehicle-mounted device.

[0125] Terminals can include various handheld devices, in-vehicle devices, wearable devices, computing devices, or other processing devices connected to a wireless modem with wireless communication capabilities; they can also include subscriber units, cellular phones, smartphones, wireless data cards, personal digital assistant (PDA) computers, tablet computers, wireless modems, handheld devices, laptop computers, cordless phones, wireless local loop (WLL) stations, machine-type communication (MTC) terminals, user equipment (UE), mobile stations (MS), terminal devices, or relay user equipment, etc. Relay user equipment can be, for example, a 5G residential gateway (RG). For ease of description, the user equipment mentioned above is collectively referred to as a terminal in this application.

[0126] It should be understood that the terminal in the embodiments of this application can also be a terminal in various vertical industry application fields such as Internet of Things terminal devices, ports, smart factories, railway transportation, logistics, drones, and autonomous vehicles. For example, mobile robots, automated guided vehicles (AGVs), autonomous vehicles, control equipment and sensors on trains, and control equipment and sensors deployed in factories.

[0127] The system also includes a network exposure function (NEF) element, which is mainly used to support the exposure of capabilities and events.

[0128] Optionally, the system may also include third-party agent networks, such as 3rd Agent (MM LLM4) and 3rd Agent (MM LLM5).

[0129] This application also involves some network elements, such as access and mobility management function (AMF) network elements, session management function (SMF) network elements, and unified data function (UDM) network elements.

[0130] UDM network elements are mainly responsible for user contract management, access authorization, and authentication information generation.

[0131] The AMF (Agency Default Manager) network element primarily performs functions such as mobility management, access authentication, and authorization. It is also responsible for transmitting user policies between the terminal and the PCF (Programmable Default Manager) network element. The SMF (Session Default Manager) network element is mainly responsible for session management in the network architecture. Its functions include session establishment, modification, and release. For example, the session management network element assigns IP addresses to terminals or selects a UPF (User-Defined Default Manager) to provide packet forwarding functionality.

[0132] SMF network elements are mainly used for session management.

[0133] It should be noted that the AN, AMF, SMF, UDM, and UPF network elements mentioned in the embodiments of this application are merely names and do not limit the device itself. In 5G networks and other future networks, the network elements or entities corresponding to AN, AMF, SMF, UDM, and UPF network elements may also have other names, and the embodiments of this application do not impose specific limitations on them.

[0134] In this application embodiment, the specific structure of the execution subject of the communication method of an intelligent agent network is not particularly limited, as long as communication can be performed according to the communication method of the intelligent agent network of this application by running a program that records the code of the communication method of the intelligent agent network of this application. For example, the execution subject of the communication method of the intelligent agent network provided in this application embodiment can be a functional module in the artificial intelligence agent function network element that can call and execute a program, or a communication device applied in the artificial intelligence agent function network element, such as a chip, chip system, integrated circuit, etc. These chips, chip systems, and integrated circuits can be set inside the artificial intelligence agent function network element, or they can be independent of the artificial intelligence agent function network element, and this application embodiment does not impose any restrictions. The execution subject of the communication method of the intelligent agent network provided in this application embodiment can be a functional module in the intelligent agent registration function network element that can call and execute a program, or a communication device applied in the intelligent agent registration function network element, such as a chip, chip system, integrated circuit, etc. These chips, chip systems, and integrated circuits can be set inside the intelligent agent registration function network element, or they can be independent of the intelligent agent registration function network element, and this application embodiment does not impose any restrictions. The following embodiments are described using two examples: one where the execution subject of a communication method in an intelligent agent network is an artificial intelligence agent function network element, and the other where the execution subject is an intelligent agent registration function network element. Unless otherwise specified, the solutions in the following embodiments can be combined.

[0135] As shown in Figure 5, Figure 5 illustrates a communication method for an intelligent agent network provided in an embodiment of this application. The method includes:

[0136] Step 501: The intelligent agent instance corresponding to the first terminal receives the instruction information from the first terminal. Accordingly, the first terminal sends the instruction information to the intelligent agent instance corresponding to it. The instruction information is used to indicate the user's intent. The first terminal is associated with the user.

[0137] In one possible embodiment, the indication information includes the identifier of the first terminal.

[0138] In this context, the user's intent represents the task that the user requires the intelligent agent network to perform.

[0139] For example, if a user's intention is to find a lost item, the intelligent agent instance corresponding to the first terminal can determine, based on the intention, that the tasks the first terminal needs to perform are locating the lost item and retrieving it. The user's intention can be forwarded by the first terminal after receiving the user's intention, or it can be sent by the first terminal based on the user's previous settings.

[0140] In this context, the association between the first terminal and the user indicates that the first terminal belongs to the terminals registered by the user. In other words, the first terminal is among the terminals registered by the user. The user's intent can be sent to the corresponding agent instance through the first terminal.

[0141] In one possible implementation, the first terminal sends indication information to an intelligent session function (ISF) network element. The ISF network element determines the intelligent agent instance corresponding to the first terminal based on the identifier of the first terminal in the indication information. The ISF network element then sends indication information to the intelligent agent instance corresponding to the first terminal.

[0142] The instruction information sent by the first terminal to the ISF network element can be the same as or different from the instruction information sent by the ISF network element to the intelligent agent instance corresponding to the first terminal.

[0143] Step 502: The intelligent agent instance corresponding to the first terminal determines the target intelligent agent instance according to the instruction information.

[0144] The capabilities of the target intelligent agent instance include the capabilities of the terminal corresponding to the target intelligent agent instance and the capabilities granted to the terminal by the network. These capabilities are used to execute tasks that achieve the intended purpose.

[0145] Among these, the terminal's capabilities refer to its basic capabilities. Basic capabilities can be understood as those inherent to the terminal itself, not requiring network provision. Examples include the terminal's device type, manufacturer, model, and remaining battery life. Different types of terminals also possess specific capabilities.

[0146] For example, when the terminal is a drone, its capabilities may also include action command sequences (e.g., takeoff, landing, forward, backward, left turn, right turn), audio and video transmission, etc. When the terminal is a robot, its capabilities may also include fixed functions, such as cleaning a room. When the terminal is a robot dog, its capabilities include action command sequences (e.g., forward, backward, left turn, right turn, stand up, lie down, sit down), and extended attachments (e.g., LiDAR, robotic arm).

[0147] The capabilities that the network provides to the terminal are referred to as network empowerment. Network empowerment can be understood as the network enhancing the terminal's capabilities through network technologies (such as infrastructure construction, data transmission, service integration, and resource virtualization).

[0148] In one possible implementation, the network empowers terminals in ways including but not limited to: resource sharing, which means the network allows terminals to access shared resources and services, such as cloud computing resources, databases, and storage space; information access, which means terminals can access large amounts of information and data through the network, such as web pages, online services, databases, and knowledge bases; communication capabilities, which means the network provides communication capabilities between terminals and between terminals and servers, supporting data exchange and message passing; and remote access and control, which means the network enables users to remotely access and manage terminal devices, enabling remote work, study, and control.

[0149] For example, the terminal is a smartphone, and the capabilities that the network provides to the smartphone can include: multimodal conversion (e.g., text-to-speech, speech-to-text, image-to-text, text-to-image, audio-to-video, video-to-text, image-to-audio, text-to-video, etc.), application programming interface calls, and multi-agent group management. For instance, a smartphone can access diverse data sources from the network (including text, images, audio, and video), and then utilize cloud computing resources and storage capabilities to process and transform different types of data, thereby enabling the network to provide capabilities to the smartphone.

[0150] For example, if the terminal is a drone, the capabilities that the network provides to the drone can include: dynamic path planning, beyond-line-of-sight obstacle avoidance, multimodal commands, and real-time voice control. Taking dynamic path planning as an example, the network can provide the drone with real-time traffic information, geographic data, and other relevant data. The drone, by leveraging the network's resource sharing and accessing cloud computing resources, runs complex path planning algorithms, processes large amounts of data, and quickly derives the optimal path, thus enabling the network to endow the drone with dynamic path planning capabilities.

[0151] For example, the terminal is a robot, and the capabilities the network provides to the robot can include cooking. For instance, the robot can access various recipes and cooking tutorials online, learning different cooking methods and ingredient pairings. It can also analyze online videos using cloud computing resources, learning the movements and techniques of human chefs and imitating the cooking process, thus enabling the network to endow the robot with cooking abilities.

[0152] For example, if the terminal is a digital human, the capabilities that the network provides to the digital human can include: trip planning, such as planning trips based on the user's schedule. For instance, the digital human can access a large amount of geographic, transportation, and timetable data through the network, and can use this data for trip planning.

[0153] It is worth noting that some terminals support local intelligent agents, and the capabilities of such terminals may include network-enabled capabilities that are applicable to other terminals.

[0154] For example, consider a smart car as the terminal. Since smart cars support local intelligent agents, their capabilities can include dynamic path planning, line-of-sight obstacle avoidance, and multimodal commands. Dynamic path planning for drones is a capability provided by the network, while multimodal commands for smartphones are also network-provided capabilities. The network-provided capabilities for smart cars can include beyond-line-of-sight perception. For instance, smart cars can use vehicle-to-everything (V2X) communication to receive information from other vehicles, road infrastructure, and pedestrians, enabling beyond-line-of-sight perception.

[0155] In one possible embodiment of this application, the capabilities that the network provides to the terminal can be planned and combined.

[0156] For example, the terminal is a drone, and the network provides the drone with the capability of dynamic path planning. Dynamic path planning can combine the drone's basic capabilities, such as when to control the drone to ascend, turn, descend, and take pictures.

[0157] For example, the terminal is a smart car, and the network provides the smart car with the capability of beyond-line-of-sight perception. Beyond-line-of-sight perception can plan the smart car's own capabilities, such as autonomous driving and braking, to make turns in advance and avoid obstacles.

[0158] For example, the terminal is a nanny robot, and the network provides the nanny robot with the ability to cook. Cooking can be planned based on the nanny robot's own capabilities, such as obtaining recipes from the network and the capabilities of its robotic arm, so that the nanny robot can complete the cooking.

[0159] In one possible embodiment of this application, step 502 described above can be implemented by the following method:

[0160] Step 5021a: The agent instance corresponding to the first terminal determines the task based on the intent.

[0161] The task of achieving the intention can be broken down into multiple sub-tasks.

[0162] For example, if the user's intention is to find a lost item, the intelligent agent instance corresponding to the first terminal determines the task that the first terminal needs to perform based on the intention: to locate and retrieve the item. This task can be divided into two sub-tasks: Task 1 (locating the lost item) and Task 2 (retrieving the lost item).

[0163] Step 5021b: Determine the agent instance corresponding to the first terminal. The target agent instance includes agent instances with the ability to perform tasks.

[0164] The target agent instance can include multiple agent instances.

[0165] For example, the target agent instances corresponding to the first terminal include agent instance 1, agent instance 2, and agent instance 3. Among them, agent instance 1 can be used to execute subtask 1, agent instance 2 can be used to execute subtask 2, and agent instance 3 can be used to execute subtask 3.

[0166] In one possible embodiment of this application, the agent instance corresponding to the first terminal can determine that the target agent instance includes an agent instance with the ability to perform tasks, based on the capabilities required by the decomposed sub-tasks. The specific method includes:

[0167] Step 1: Determine the ability of the agent instance corresponding to the first terminal to execute the first subtask.

[0168] For example, if the subtask 1 is to search for items in a park, then the agent instance corresponding to the first terminal determines that the capability required to perform subtask 1 is capability 1 (e.g., flight capability, image acquisition capability). If the subtask 2 is to retrieve items, then the agent instance corresponding to the first terminal determines that the capability required to perform subtask 2 is capability 2 (e.g., movement capability, gripping capability, path planning capability).

[0169] Step 2: Determine the target intelligent agent instance corresponding to the first terminal. The target intelligent agent instance includes the first intelligent agent instance. The first intelligent agent instance has the capability to execute the first subtask.

[0170] For example, the agent instance corresponding to the first terminal determines the agent instance with capability 1 based on capability 1. Since the target agent instance's capability includes capability 1, the agent instance corresponding to the first terminal determines that the target agent instance includes this agent instance.

[0171] In one possible embodiment of this application, the determination of the target intelligent agent instance corresponding to the first terminal, including the first intelligent agent instance, can be achieved by requesting the intelligent agent registration function network element. The specific method includes:

[0172] Step 21a: The agent instance corresponding to the first terminal sends a request message to the agent registration function network element. Correspondingly, the agent registration function network element receives the request message from the agent instance corresponding to the first terminal. The request message indicates the capability to execute the first subtask.

[0173] The agent registration function network element stores information about agent instances, including the capabilities of the terminal corresponding to the agent instance and the capabilities granted to the terminal by the network. Agent instances include target agent instances.

[0174] The request message includes information about the required capabilities.

[0175] For example, the agent instance corresponding to the first terminal determines that performing the task to achieve the intention requires capability 1 and capability 2. The agent instance corresponding to the first terminal sends the information of capability 1 and capability 2 to the agent registration function network element. After receiving the information of capability 1 and capability 2, the agent registration function network element determines the agent instance with capability 1 and the agent instance with capability 2.

[0176] Step 21b: The agent registration function network element sends the information of the first agent instance to the agent instance corresponding to the first terminal. Correspondingly, the agent instance corresponding to the first terminal receives the information from the first agent instance of the agent registration function network element.

[0177] The information of the first intelligent agent instance includes its address information. The first intelligent agent instance is the target intelligent agent instance for executing the task. The intelligent agent instance corresponding to the first terminal determines the first intelligent agent instance from the artificial intelligence agent function network element based on its address information.

[0178] In another possible embodiment of this application, the agent instance corresponding to the first terminal is associated with the first agent instance, and the determination of the target agent instance, including the first agent instance, can be directly achieved through the agent instance corresponding to the first terminal. Specific methods include:

[0179] Step 22a: The first agent instance corresponding to the first terminal is determined from at least one agent instance associated with the agent instance corresponding to the first terminal.

[0180] One possible implementation of associating the target agent instance with the agent instance corresponding to the first terminal is that the target agent instance and the agent instance corresponding to the first terminal are located in the same user-corresponding agent instance.

[0181] For example, if user A is associated with a first terminal and a second terminal, when the first terminal and the second terminal create corresponding intelligent agent instances, the AI ​​agent function network element determines that both the first terminal and the second terminal belong to user A. Then, it creates a corresponding intelligent agent instance for user A. The intelligent agent instance corresponding to the first terminal is in the corresponding intelligent agent instance of user A, and the intelligent agent instance corresponding to the second terminal is also in the corresponding intelligent agent instance of user A.

[0182] For example, the first terminal is a smartphone (UE1), and the second terminal is a drone (UE2). UE1 and UE2 belong to user A. After UE1 creates a corresponding agent instance (Agent1a), it determines that UE2 also needs to create a corresponding agent instance. Therefore, it creates a corresponding agent instance (Agent1) for user A, with Agent1a contained within Agent1. UE2 creates a corresponding agent instance (Agent1b), which is also contained within Agent1. Agent1b is the target agent instance.

[0183] In one possible embodiment, at least one smart agent instance associated with the smart agent instance corresponding to the first terminal includes: the terminal corresponding to at least one smart agent instance and the first terminal belong to the same user.

[0184] In one possible implementation, the agent instance corresponding to the first terminal determines the first agent instance from the agent instance corresponding to the user.

[0185] For example, consider a smartphone as the first terminal and a drone as the second terminal, both belonging to the same user. The agent instance (Agent1a) corresponding to the smartphone determines that the agent instance (Agent1b) corresponding to the drone belongs to the same user's agent instance (Agent1). Since Agent1b possesses the required capabilities, Agent1a identifies Agent1b as the first agent instance within Agent1.

[0186] In one possible embodiment of this application, the indication information includes an identifier of the second terminal. It is understood that the indication information already specifies the terminal used to perform the task. Step 502 above can also be implemented by the following methods, specifically including:

[0187] Step 5022a: The intelligent agent instance corresponding to the first terminal determines the target intelligent agent instance to include the intelligent agent instance corresponding to the second terminal based on the identifier of the second terminal.

[0188] For example, the instruction information includes the identifier of terminal 2. The agent instance corresponding to the first terminal determines the agent instance 2 corresponding to terminal 2 based on the identifier of terminal 2. Since agent instance 2 corresponding to terminal 2 has the capability to perform the task, the agent instance corresponding to the first terminal determines that the target agent instance includes agent instance 2.

[0189] Step 503: The agent instance corresponding to the first terminal triggers the target agent instance to execute tasks according to the capabilities of the target agent instance to achieve the intent.

[0190] In one possible embodiment, the implementation method of step 503 above includes: the intelligent agent instance corresponding to the first terminal sending information describing the first subtask to the target intelligent agent instance. Correspondingly, the target intelligent agent instance receives the information describing the first subtask from the intelligent agent instance corresponding to the first terminal.

[0191] The information used to describe the first subtask can be the intent of the first subtask, which instructs the target agent instance to invoke the capabilities granted to the terminal by the network and / or the capabilities of the terminal to execute the first subtask.

[0192] For example, after understanding the intent based on the instruction information, the agent instance corresponding to the first terminal decomposes the task into multiple first sub-tasks (e.g., task 1 and task 2). It then determines that completing task 1 requires the capabilities of agent instance 1, and completing task 2 requires the capabilities of agent instance 2. The agent instance corresponding to the first terminal sends the intent for task 1 to agent instance 1, requesting agent instance 1 to execute task 1; and sends the intent for task 2 to agent instance 2, requesting agent instance 2 to execute task 2.

[0193] This application receives instruction information from a first terminal via an AI agent function network element, indicating the user's intent. Based on the intent, it then determines at least one intelligent agent instance required to achieve it. Since each intelligent agent instance contains the capabilities of the corresponding terminal and the capabilities granted to the terminal by the network, the AI ​​agent function network element can request at least one intelligent agent instance to perform the task required to achieve the intent. Because the intelligent agent instances permanently store the terminal's capabilities and the capabilities granted to the terminal by the network, the AI ​​agent function network element can utilize the intelligent agent network to perform complex intelligent operations at any time.

[0194] In one possible embodiment of this application, the target agent instance receives information describing a first subtask. In response to the information describing the first subtask, the target agent instance's capabilities are invoked to execute the first subtask.

[0195] For example, agent instance 1 is the target agent instance. After receiving the intent of task 1, agent instance 1 responds to the intent of task 1 by calling the capabilities of agent instance 1 to execute task 1.

[0196] It's worth noting that the target agent instance can also be the agent instance corresponding to the first terminal. That is, after understanding the intent, the agent instance corresponding to the first terminal decomposes the task into a first subtask, and the capabilities required by the first subtask are those of the agent instance corresponding to the first terminal. The specific method described above is as follows: the agent instance corresponding to the first terminal receives information describing the first subtask, and in response to this information, invokes the capabilities of the agent instance corresponding to the first terminal to execute the first subtask.

[0197] In one possible embodiment of this application, the specific method for the target intelligent agent instance to invoke the capabilities of the target intelligent agent instance to execute the first sub-task includes: the target intelligent agent instance planning and combining the capabilities of the terminal according to the capabilities granted to the terminal by the network to execute the first sub-task.

[0198] For specific implementation details, please refer to the above embodiments, which will not be repeated here.

[0199] The above embodiment describes how the intelligent agent instance corresponding to the first terminal invokes the capabilities of intelligent agent instances corresponding to other terminals to perform tasks and achieve the user's intent. Prior to this, the terminals (including the first terminal and other terminals) need to create corresponding intelligent agent instances within the artificial intelligence agent function network element. As shown in Figure 6, Figure 6 illustrates a communication method for an intelligent agent network provided in this application embodiment, used to create corresponding intelligent agent instances for each terminal within the artificial intelligence agent function network element. The method includes:

[0200] Step 601: The AI ​​agent function network element creates an intelligent agent instance corresponding to the first terminal.

[0201] The first terminal can be the same as the first terminal in the above embodiments, or it can be a different terminal.

[0202] There can be one or more AI agent function network elements, and each AI agent function network element creates one or more intelligent agent instances corresponding to terminals. For example, AI agent function network element 1 determines the intelligent agent instances corresponding to terminal 1 and terminal 4 respectively, and AI agent function network element 2 determines the intelligent agent instances corresponding to terminal 2 and terminal 3 respectively.

[0203] In one possible embodiment of this application, prior to step 601 described above, the method provided in this application further includes:

[0204] Step 600: The AI ​​agent function network element receives information from the first terminal indicating the capabilities of the first terminal.

[0205] For example, the first terminal sends a request message to the AI ​​agent function network element. The request message includes information indicating the capabilities of the first terminal. Optionally, the request message may also include the identifier of the first terminal.

[0206] In one possible implementation, the AI ​​agent function network element can directly obtain information from the first terminal that indicates the capabilities of the first terminal.

[0207] For example, if the first terminal supports the new non-access stratum (NAS), the first terminal can directly send information indicating the capabilities of the first terminal to the AI ​​agent function network element.

[0208] In another possible implementation, the AI ​​agent function network element receives information from the control plane of the core network to indicate the capabilities of the first terminal.

[0209] For example, if the first terminal does not support the new non-access stratum (NAS), then the first terminal needs to first send information indicating the capabilities of the first terminal to the control plane of the core network, and then the control plane of the core network sends information indicating the capabilities of the first terminal to the AI ​​agent function network element.

[0210] The specific implementation of step 601 above can be achieved by using an artificial intelligence agent function network element to create an intelligent agent instance corresponding to the first terminal based on the capabilities of the first terminal.

[0211] In one possible embodiment of this application, the specific implementation method of the AI ​​agent function network element creating an intelligent agent instance corresponding to the first terminal based on the capabilities of the first terminal includes:

[0212] Step 1: The AI ​​agent function network element determines the intelligent agent template based on the capabilities of the first terminal and / or the contract information of the first terminal.

[0213] The agent template is used to provide a basic framework for agent instances, enabling AI agent function network elements to create agent instances based on the capabilities of the first terminal.

[0214] In one possible implementation, the agent template is stored in the Agent Template Repository (ATR), and the AI ​​agent function network element retrieves the corresponding agent template from the ATR based on the capabilities of the first terminal and / or the contract information of the first terminal.

[0215] The agent template in the ATR can be provided by a third party or configured by operations, administration, and maintenance (OAM). For example, a third party or OAM can send an agent template to the AI ​​agent function network element, which will then save the agent template in the ATR after confirmation.

[0216] In another possible implementation, the AI ​​agent function network element interacts with the LLM to obtain the agent template.

[0217] Step 2: The AI ​​agent function network element determines the intelligent agent instance corresponding to the first terminal based on the intelligent agent template.

[0218] Step 602: The AI ​​agent function network element instructs the agent instance to send a registration request message to the agent registration function network element. Correspondingly, the agent registration function network element receives the registration request message from the agent instance.

[0219] The registration request message includes the capabilities of the agent instance. These capabilities include the capabilities of the first terminal and the capabilities granted to the first terminal by the network.

[0220] Optionally, the registration request message may also include the identifier of the first terminal and the identifier of the first AI agent function network element. The first AI agent function network element is the AI ​​agent function network element where the intelligent agent instance corresponding to the first terminal resides.

[0221] In one possible embodiment of this application, multiple smart agent instances corresponding to terminals can be associated. Specific methods include:

[0222] Step 1: The AI ​​agent function network element receives information from the second terminal indicating the capabilities of the second terminal.

[0223] For example, the second terminal sends information about its capabilities to the AI ​​agent function network.

[0224] Step 2: If the first terminal and the second terminal belong to the same user, the AI ​​agent function network element creates an intelligent agent instance corresponding to the user. The capabilities of the intelligent agent instance corresponding to the user include the capabilities of the intelligent agent instance corresponding to the first terminal and the capabilities of the intelligent agent instance corresponding to the second terminal.

[0225] Among them, the AI ​​agent function network element determines that the first terminal and the second terminal belong to the same user, which can be determined based on the identifier of the first terminal and the identifier of the second terminal.

[0226] For example, if both the identifier of the first terminal and the identifier of the second terminal include the identifier of user A, then it can be determined that the first terminal and the second terminal belong to user A.

[0227] Specifically, the AI ​​agent function network element creates a corresponding intelligent agent instance for the user based on information used to instruct the capabilities of the first terminal. The user-corresponding intelligent agent instance includes the intelligent agent instance corresponding to the first terminal.

[0228] For example, based on the capabilities of the first terminal (UE1), a user-corresponding agent instance (Agent1) is created, and the agent instance (Agent1a) corresponding to UE1 is contained in Agent1.

[0229] When the first terminal and the second terminal belong to the same user, the method provided in this application embodiment further includes: the artificial intelligence agent function network element instructing the intelligent agent instance corresponding to the user to send a second request message to the intelligent agent registration function network element. Correspondingly, the intelligent agent registration function network element receives the second request message from the intelligent agent instance corresponding to the user.

[0230] The second request message is used to indicate the capabilities of the intelligent agent instance corresponding to the user. The capabilities of the intelligent agent instance corresponding to the user include the capabilities of the intelligent agent instance corresponding to the first terminal and the capabilities of the intelligent agent instance corresponding to the second terminal.

[0231] For example, the second request message may include the capabilities of the intelligent agent instance corresponding to the first terminal and the capabilities of the intelligent agent instance corresponding to the second terminal. Optionally, the second request message may also include the identifier of the first terminal, the identifier of the second terminal, the identifier of the first artificial intelligence agent function network element, and the identifier of the intelligent agent instance corresponding to the user.

[0232] In one possible implementation, when the third terminal, the first terminal, and the second terminal all belong to the same user, after the third terminal creates a corresponding agent instance, the user's corresponding agent instance is updated, and a third request message is sent to the agent registration function network element. The third request message is used to instruct the agent registration function network element to update the user's corresponding agent instance.

[0233] The third request message includes the capabilities of the intelligent agent instance corresponding to the third terminal.

[0234] For example, the agent instance (Agent1) corresponding to user A includes the agent instance (Agent1a) corresponding to the first terminal and the agent instance (Agent1b) corresponding to the second terminal. The third terminal creates a corresponding agent instance (Agent1c) and associates it with Agent1. Agent1 sends a third request message to the agent registration function network element. The third request message includes the capabilities of Agent1c.

[0235] The following describes a possible embodiment of an AI agent function network element creating agent instances for multiple terminals, and one of the multiple terminals triggering a target agent instance to perform a task based on the capabilities of the target agent instance to achieve an intention.

[0236] As shown in Figure 7, Figure 7 is a schematic diagram of a terminal creating a corresponding intelligent agent instance according to an embodiment of this application. The terminal includes terminal 1, terminal 2, terminal 3, and terminal 4.

[0237] Step 1: The terminal sends a first registration request message to the AI ​​agent function network element. Correspondingly, the AI ​​agent function network element receives the first registration request message from the terminal. The first registration request message is used to indicate the terminal's capabilities.

[0238] Among them, there can be one or more AI agent function network elements.

[0239] For example, referring to Figure 7, taking the AI ​​agent function network elements including AIAF1 and AIAF2 as an example, the implementation method of step 1 above includes:

[0240] Step 1a: Terminal 1 sends a first registration request message to AIAF1. Correspondingly, AIAF1 receives the first registration request message from Terminal 1. The first registration request message includes the capabilities of Terminal 1.

[0241] For example, terminal 1 is a smartphone that supports the new NAS. The smartphone sends a first registration request message to AIAF1. The first registration request message includes the device type (smartphone), manufacturer (xx), model (xx), remaining battery life (xx hours), etc.

[0242] Step 1b: Terminal 2 sends a first registration request message to AIAF2. Correspondingly, AIAF2 receives the first registration request message from Terminal 2. The first registration request message includes the capabilities of Terminal 2.

[0243] For example, terminal 2 is a drone that supports the new NAS. The drone sends a first registration request message to AIAF2. The first registration request message includes the device type (drone), manufacturer (xx), model (xx), remaining flight time (xx hours), real-time audio and video transmission, and action command sequence (take-off, landing, forward, backward, left turn, right turn), etc.

[0244] Step 1c: Terminal 3 sends a first registration request message to AIAF2. Correspondingly, AIAF2 receives the first registration request message from terminal 3. The first registration request message includes the capabilities of terminal 3.

[0245] For example, terminal 3 is a smart car that supports NAS. The smart car sends a first registration request message to AIAF2. The first registration request message includes device type (smart car), manufacturer (xx), model (xx), remaining battery life (xx hours), location (latitude and longitude), real-time audio and video transmission, dynamic path planning, obstacle avoidance within line of sight, multimodal commands, etc.

[0246] Step 1d: Terminal 4 sends a first registration request message. The first registration request message includes the capabilities of Terminal 4.

[0247] For example, terminal 4 is a nanny robot and does not support the new NAS. The nanny robot first sends an access request message to the core network control plane, and the control plane sends a first registration request message to AIAF1 based on the access request message. The first registration request message includes the device type (nanny robot), manufacturer (xx), model (xx), remaining battery life (xx hours), real-time video transmission, room cleaning, etc.

[0248] Step 2: The AI ​​agent function network element instantiates the intelligent agent based on the terminal's capabilities and contract information.

[0249] Among them, the terminal's contract information is used to indicate whether the terminal has signed up for intelligent agent access, the services it is empowered with, and the level of intelligent empowerment it can receive.

[0250] In one possible implementation, the AI ​​agent function network element determines the agent template based on the terminal's capabilities and contract information, and creates an agent instance corresponding to the terminal based on the agent template.

[0251] In one possible implementation, as shown in Figure 8, which is a schematic diagram of a process for a third party to provide an agent template to an agent network according to an embodiment of this application, the method includes:

[0252] Step a: The application function (AF) network element sends an agent instance template provision message to the network exposure function (NEF) network element. Correspondingly, the NEF network element receives the agent instance template provision message from the AF network element.

[0253] Among them, the agent template provides the ability to include agent template identifier, agent instance, terminal information and agent instance in the message.

[0254] For example, agent instances include, but are not limited to: agent executors, vector databases, large models, and tools; terminal information includes, but is not limited to: terminal type, manufacturer, and model.

[0255] Step b: The NEF network element provides messages to the agent instance template for authentication and authorization.

[0256] Step c: The NEF network element sends an agent instance template provision message to the AIAF grid operator. Correspondingly, the AIAF network element receives the agent instance template provision message from the NEF network element.

[0257] Step d: The AIAF network element saves the agent instance template to the ATR and sends an acknowledgment message to the NEF network element. Correspondingly, the NEF network element receives the acknowledgment message from the AIAF network element.

[0258] Step e: The NEF network element sends an acknowledgment message to the AF network element. Correspondingly, the AF network element receives the acknowledgment message from the NEF network element.

[0259] In another possible implementation, as shown in Figure 9, which is a flowchart illustrating an OAM configuration agent template provided in an embodiment of this application, the method includes:

[0260] Step a: OAM sends an agent template provision message to the AIAF network element. Correspondingly, the AIAF network element receives the agent template provision message from OAM.

[0261] Among them, the agent template provides the ability to include agent template identifier, agent instance, terminal information and agent instance in the message.

[0262] Step b: The AIAF network element saves the agent instance template to the ATR.

[0263] Step c: The AIAF network element sends an acknowledgment message to the OAM. Correspondingly, the OAM receives the acknowledgment message from the AIAF network element.

[0264] For example, referring to Figure 7, the method for instantiating intelligent agents by artificial intelligence agent function network elements includes:

[0265] Step 2a: AIAF1 creates an agent instance Agent1 corresponding to terminal 1 based on the capabilities of terminal 1 and the contract information.

[0266] Step 2b: AIAF2 creates an agent instance Agent2 corresponding to terminal 2 based on the capabilities of terminal 2 and the contract information.

[0267] Step 2c: AIAF2 creates an agent instance Agent3 corresponding to terminal 3 based on the capabilities of terminal 3 and the contract information.

[0268] Step 2d: AIAF1 creates the Agent4 instance corresponding to Terminal 4 based on the capabilities of Terminal 4 and the contract information.

[0269] Step 3: The agent instance corresponding to the terminal sends a second registration request message to the agent registration function network element. Correspondingly, the agent registration function network element receives the second registration request message from the agent instance corresponding to the terminal.

[0270] The second registration request message indicates the capabilities and identifier of the agent instance corresponding to the terminal. The capabilities of the agent instance include the terminal's capabilities and network capabilities. The identifier of the agent instance is used to identify the agent instance.

[0271] For example, referring to Figure 7, the method for the intelligent agent instance corresponding to the terminal to send a second registration request message to the intelligent agent registration function network element includes:

[0272] Step 3a: Agent 1, corresponding to Terminal 1, sends a second registration request message to the agent registration function network element. Correspondingly, the agent registration function network element receives the second registration request message from Agent 1.

[0273] The second registration request message includes Agent1's capabilities and its identifier. Agent1's capabilities include terminal capabilities and network empowerment. Network empowerment includes, but is not limited to, multimodal transformation (e.g., multimodal understanding, multimodal generation), application programming interface (API) calls (e.g., tool lists), and multi-agent group management. Optionally, API calls include configuration for calling third-party agent instances. Agent1's identifier is a Uniform Resource Identifier (URI) including {apiRoot1} / aiaf{agentId1}, where apiRoot1 identifies the AI ​​agent function network element, and agentId1 identifies Agent1.

[0274] Step 3b: Agent 2, corresponding to Terminal 2, sends a second registration request message to the agent registration function network element. Correspondingly, the agent registration function network element receives the second registration request message from Agent 2.

[0275] The second registration request message includes Agent2's capabilities and its identification information. Agent2's capabilities include terminal capabilities and network enablement, including but not limited to: dynamic path planning, beyond-line-of-sight obstacle avoidance (e.g., 500 meters), and multimodal commands (e.g., real-time voice control). Agent2's identification information is a URI, including {apiRoot2} / aiaf{agentId2}, where apiRoot2 identifies the AI ​​agent function network element, and agentId2 identifies Agent2.

[0276] Step 3c: Agent 3, corresponding to terminal 3, sends a second registration request message to the agent registration function network element. Correspondingly, the agent registration function network element receives the second registration request message from Agent 3.

[0277] The second registration request message includes: Agent3's capabilities and Agent3's identification information. Agent3's capabilities include terminal capabilities and network empowerment, with network empowerment including but not limited to: beyond-line-of-sight perception. Agent3's identification information is a URI, including {apiRoot3} / aiaf{agentId3}, where apiRoot3 identifies the AI ​​agent function network element, and agentId3 identifies Agent3.

[0278] In step 3d, Agent 4, corresponding to terminal 4, sends a second registration request message to the agent registration function network element. Correspondingly, the agent registration function network element receives the second registration request message from Agent 4.

[0279] The second registration request message includes Agent4's capabilities and its identification information. Agent4's capabilities include terminal capabilities and network empowerment, with network empowerment including but not limited to: cooking. Agent4's identification information is a URI, including {apiRoot4} / aiaf{agentId4}, where apiRoot4 identifies the AI ​​agent function network element, and agentId4 identifies Agent4.

[0280] Based on the embodiment shown in Figure 7, the method provided in this application further includes: the intelligent agent instance corresponding to the first terminal executes a task through the target intelligent agent instance to achieve the intent initiated by the first terminal. As shown in Figure 10, taking the first terminal as terminal 1 as an example, after step 3 above, the method includes:

[0281] Step 1: Terminal 1 sends an intent to Agent 1. Agent 1 then receives the intent from Terminal 1.

[0282] The intent is used to instruct terminal 1 on the task to be performed.

[0283] Step 2: Agent1 understands the intent of Terminal 1 and breaks down the task that Terminal 1 needs to perform into multiple sub-tasks.

[0284] For example, Agent1 breaks down the task that terminal 1 needs to perform into subtask 1, subtask 2, subtask 3 and subtask 4.

[0285] Step 3: Agent1 obtains the address information of the target intelligent agent instance.

[0286] In one possible implementation, Agent1 determines the capabilities required to execute each subtask based on multiple subtasks. For example, subtask 1 requires capability 1, subtask 2 requires capability 2, subtask 3 requires capability 3, and subtask 4 requires capability 4.

[0287] Agent1 sends the capabilities required for each subtask to the ARF network element. Correspondingly, the ARF network element receives the capabilities required for each subtask from Agent1. Based on the capabilities required for each subtask, the ARF network element determines the agent instance with the capability. The ARF network element then sends the address information of the agent instance with the capability to Agent1.

[0288] For example, the ARF network element determines that the intelligent agent instance with capability 1 is Agent2, the intelligent agent instance with capability 2 is Agent3, the intelligent agent instance with capability 3 is Agent5, and the intelligent agent instance with capability 4 is Agent4.

[0289] Step 4: Agent1 sends information to the target agent instance to indicate the subtask.

[0290] For example, Agent1 sends the intent of subtask 1 to Agent2; Agent1 sends the intent of subtask 2 to Agent3; Agent1 sends the intent of subtask 3 to Agent5; Agent1 sends the intent of subtask 4 to Agent4.

[0291] Step 5: The target agent instance triggers the execution of the subtask by the target agent instance and the terminal corresponding to the target agent instance based on the information used to indicate the subtask.

[0292] In one possible implementation, the terminal corresponding to the target agent instance does not support agents. The target agent instance determines the instruction information of the terminal corresponding to it based on the information used to indicate the subtask. The instruction information is used to instruct the terminal corresponding to the target agent instance to execute the subtask.

[0293] For example, in step 5a, Agent2, based on the intent of subtask 1, is triggered to invoke network empowerment, such as wireless sensing capabilities, allocate computing resources, and establish a quality of service flow. Agent2 performs instruction escaping based on the intent of subtask 1, sending the instructions executed by terminal 2 corresponding to Agent2 to terminal 2, so that terminal 2 executes subtask 1.

[0294] For example, in step 5b, Agent4, based on the intent of subtask 4, triggers Agent4 to invoke network empowerment, such as wireless sensing capabilities, allocate computing resources, and establish a quality of service flow. Agent4 performs instruction escaping based on the intent of subtask 4, sending the instructions executed by terminal 4 corresponding to Agent4 to terminal 4, so that terminal 4 executes subtask 4.

[0295] In another possible implementation, the terminal corresponding to the target agent instance supports agents, and the target agent instance sends information to the terminal corresponding to the target agent instance to indicate subtasks.

[0296] For example, in step 5c, Agent 3 sends the intent of subtask 2 to terminal 3. Agent 3 invokes the capabilities and network empowerment of terminal 3, such as beyond-line-of-sight perception, so that terminal 3 executes subtask 2.

[0297] In another possible implementation, the target agent instance instructs the terminal corresponding to the target agent instance to execute the subtask based on the information used to instruct the subtask.

[0298] For example, Agent5, configured by a third party, instructs the terminal corresponding to Agent5 to execute subtask 3 according to the intent of subtask 3.

[0299] The following describes the specific implementation method of the AI ​​agent function network element for creating intelligent agent instances for multiple terminals:

[0300] As shown in Figure 11, Figure 11 is a schematic diagram illustrating a specific implementation method for a terminal to create a corresponding intelligent agent instance according to an embodiment of this application. The terminal supports the non-access layer. The specific method is as follows:

[0301] Step 1101: The terminal (UE) interacts with the Access and Mobility Management Function (AMF) network element to initiate the registration of intelligent services.

[0302] Optionally, during the process of UE initiating registration for intelligent services, UE and AMF network element negotiate the capability to support intelligent sessions. If the negotiation function is successful, AMF network element returns the address of ISF network element to UE.

[0303] Step 1102: The UE interacts with the Session Management Function (SMF) network element or the User Plane Function (UPF) network element to establish a PDU session.

[0304] Optionally, during the process of the UE establishing a protocol SATA unit (PDU) session, the UE and the SMF network element negotiate the capability to support intelligent sessions. If the negotiation function is successful, the SMF network element returns the address of the ISF network element to the UE.

[0305] Step 1103: The UE sends the terminal's capabilities and terminal identifier to the ISF network element. Correspondingly, the ISF network element receives the terminal's capabilities and terminal identifier from the UE.

[0306] Optionally, if the UE supports local intelligent agents, the UE can also send local intelligent agent capabilities to the ISF network element.

[0307] Step 1104: ISF network element determines AIAF network element.

[0308] For example, among multiple AIAF network elements, the ISF network element determines AIAF1 to create an intelligent agent instance for the UE.

[0309] Step 1105: The ISF network element sends a first request message to the AIAF network element. Correspondingly, the AIAF network element receives the first request message from the ISF network element. The first request message is used to request the creation of an agent instance.

[0310] The first request message includes the terminal's capabilities and the terminal's identifier.

[0311] Optionally, if the UE supports local agents, the first request message may also include local agent capabilities.

[0312] Step 1106: The AIAF network element sends a second request message to the unified data management (UDM) network element. Correspondingly, the UDM network element receives the second request message from the AIAF network element. The second request message is used to request subscription information.

[0313] The contract information is used to indicate whether the terminal has signed up for intelligent agent access, the services it is empowered with, and the level of intelligent empowerment it can receive.

[0314] Step 1107: The UDM network element sends subscription information to the AIAF network element. Correspondingly, the AIAF network element receives the subscription information from the UDM network element.

[0315] Step 1108: The AIAF network element determines the intelligent agent template based on the terminal capabilities and contract information, and creates an intelligent agent instance based on the intelligent agent template.

[0316] In one possible implementation, the AIAF network element obtains the agent template from the ATR. The ATR can be integrated into the AIAF network element or operate independently of it; this is not limited in the embodiments of this application.

[0317] In another possible implementation, the AIAF network element obtains the agent template from the LLM.

[0318] Among them, the agent instance supports terminal capabilities and network empowerment. Network empowerment can include, but is not limited to, dynamic path planning, advanced action commands, network awareness capabilities, and the ability to call application programming interfaces.

[0319] Step 1109: The AIAF network element sends a registration request message to the ARF network element. Correspondingly, the ARF network element receives the registration request message from the AIAF network element.

[0320] The registration request message includes the capabilities and identifier of the smart agent instance corresponding to the UE.

[0321] Step 1110: The AIAF network element sends a registration response message to the ISF network element. Correspondingly, the ISF network element receives the registration response message from the AIAF network element.

[0322] The registration response message includes the capabilities of the smart agent instance corresponding to the UE.

[0323] Step 1111: The ISF network element sends a registration response message to the UE. Correspondingly, the UE receives the registration response message from the ISF network element.

[0324] Optionally, taking a smartphone as an example, the method further includes: step 1112, if the UE is creating a smart agent instance for the first time, then the UE creates an entry corresponding to the smart agent instance in the address book.

[0325] After the agent network creates an agent instance for the terminal, the terminal can also deregister to delete the created agent instance. Based on the specific implementation of the terminal creating the corresponding agent instance shown in Figure 11, and as shown in Figure 12, which is a schematic diagram of a specific implementation of a terminal deregistering an agent instance according to an embodiment of this application, including:

[0326] Step 1201: The terminal (UE) interacts with the AMF network element to initiate the registration of intelligent services.

[0327] Step 1202: The ISF network element sends a third request message to the AIAF network element. Correspondingly, the AIAF network element receives the third request message from the ISF network element. The third request message is used to request the deletion of the agent instance.

[0328] The third request message includes the identifier of the agent instance.

[0329] Step 1203: The AIAF network element deletes the corresponding agent instance based on the third request message.

[0330] Step 1204: The AIAF network element sends an acknowledgment message to the ISF network element. Correspondingly, the ISF network element receives the acknowledgment message from the AIAF network element.

[0331] The confirmation message indicates that the agent instance in the AIAF network element has been deleted.

[0332] Step 1205: The AIAF network element sends a deregistration request message to the ARF network element. The ARF network element receives the deregistration request message from the AIAF network element.

[0333] The deregistration request message is used to deregister the agent instance information in the ARF network element. The deregistration request message includes the identifier of the agent instance.

[0334] Step 1206: The ISF network element sends a deregistration response message to the UE. Correspondingly, the UE receives the deregistration response message from the ISF network element.

[0335] The registration response message indicates that the registration of smart services has been completed.

[0336] Step 1207: The UE sends a session termination request message to the SMF network element. Correspondingly, the SMF network element receives the session termination request message from the UE.

[0337] The session request end message is used to request the termination of the PDU session.

[0338] Step 1208: The UE sends a deregistration request message to the AMF network element. Correspondingly, the AMF network element receives the deregistration request message from the UE.

[0339] The registration request message is used to instruct the user to register from the AMF network element.

[0340] As shown in Figure 13, Figure 13 is a schematic diagram of another specific implementation method for creating a corresponding intelligent agent instance by a terminal according to an embodiment of this application. The difference from the specific implementation method shown in Figure 10 is that the terminal does not support the non-access layer. The specific method is as follows:

[0341] Step 1301: The terminal (UE) interacts with the AMF network element to initiate the registration of intelligent services.

[0342] Step 1302: The UE interacts with the SMF network element or UPF network element to establish a PDU session.

[0343] Step 1303: The SMF network element or UPF network element sends a fourth request message to the UDM network element. Correspondingly, the UDM network element receives the fourth request message from the SMF network element or UPF network element.

[0344] The fourth request message is used to request a subscription instruction for smart services.

[0345] Step 1304: The UDM network element sends a subscription instruction to the SMF network element or UPF network element. Correspondingly, the SMF network element or UPF network element receives the subscription instruction from the UDM network element.

[0346] The signing instruction is used to instruct SMF network elements to request the creation of intelligent agent instances.

[0347] Step 1305: The SMF network element or UPF network element sends the terminal's capabilities and terminal identifier to the ISF network element. Correspondingly, the ISF network element receives the terminal's capabilities and terminal identifier from the SMF network element or UPF network element.

[0348] Steps 1306 to 1312 are the same as steps 1104 to 1110 in the above embodiments, and will not be repeated here.

[0349] After the agent network creates an agent instance for the terminal, the terminal can also deregister to delete the created agent instance. Based on the specific implementation of the terminal creating the corresponding agent instance shown in Figure 13, and as illustrated in Figure 14, which is a schematic diagram of another specific implementation of the terminal deregistering an agent instance provided in this application embodiment, including:

[0350] Step 1401: The UE interacts with network elements in the control plane to initiate PDU session termination. These network elements include AMF, SMF, and UPF network elements.

[0351] Step 1402: The SMF network element sends a fifth request message to the ISF network element. Correspondingly, the ISF network element receives the fifth request message from the SMF network element. The fifth request message is used to request the cancellation of intelligent services.

[0352] The fifth request message includes the terminal's identifier.

[0353] Steps 1403 to 1406 are the same as steps 1202 to 1205 in the above embodiments, and will not be repeated here.

[0354] Step 1407: The ISF network element sends a deregistration response message to the SMF network element or UPF network element. Correspondingly, the SMF network element or UPF network element receives the deregistration response message from the ISF network element.

[0355] The cancellation response message is used to indicate that the smart service has been cancelled.

[0356] After an intelligent agent instance is created on the terminal, the AI ​​agent function network element executes a task based on an intent initiated by the terminal. As shown in Figure 15, Figure 15 is a schematic diagram illustrating a specific implementation method of an intelligent agent instance corresponding to a first terminal executing a task through a target intelligent agent instance based on a terminal intent, according to an embodiment of this application. Taking terminal 1 (UE1) initiating an intent as an example, the specific implementation includes:

[0357] Step 1501: UE1 sends a sixth request message to the ISF network element. Correspondingly, the ISF network element receives the sixth request message from UE1.

[0358] The sixth request message includes the identifier of UE1 and the intent of UE1.

[0359] Optionally, the sixth request message may also include an identifier of the communication peer. The communication peer is the terminal that performs the task, as indicated by UE1.

[0360] Step 1502: The ISF network element determines the intelligent agent instance corresponding to UE1 based on the sixth request message.

[0361] For example, the ISF network element determines the Agent1 corresponding to UE1 based on the identifier of UE1.

[0362] Step 1503: The ISF network element sends a seventh request message to Agent1. Correspondingly, Agent1 receives the seventh request message from the ISF network element.

[0363] The seventh request message includes the identifier of UE1 and the intent of UE1.

[0364] Optionally, the seventh request message may also include the identifier of the communication peer.

[0365] Case 1) No identifier of the communication peer is indicated.

[0366] Step 1504A: Agent1 understands the intent, decomposes the task based on the understood intent, and determines the agent capabilities required to complete the sub-tasks.

[0367] For example, Agent1 determines that it needs to complete subtask 1 and subtask 2 based on the understood intent. Completing subtask 1 requires agent capability 1, and completing subtask 2 requires agent capability 2.

[0368] Step 1505A: Agent1 sends the eighth request message to the ARF network element. Correspondingly, the ARF network element receives the eighth request message from Agent1.

[0369] The eighth request message requests the address of the target agent instance. The target agent instance is an agent instance that meets the required capabilities.

[0370] Step 1506A: The ARF network element determines the address of the target agent instance based on the eighth request message.

[0371] Step 1507A: The ARF network element sends the address of the target agent instance to Agent1. Correspondingly, Agent1 receives the address of the target agent instance from the ARF network element.

[0372] For example, if Agent2 is an agent instance that satisfies Agent capability 1 and Agent3 is an agent instance that satisfies Agent capability 2, then the ARF network element sends the addresses of Agent2 and Agent3 to Agent1.

[0373] Taking Agent2 and Agent3 as target agent instances as an example, Agent2 does not support local agents, while Agent3 does. The method also includes:

[0374] Step 1508A: Agent1 sends the intent of Subtask 1 to Agent2. Correspondingly, Agent2 receives the intent of Subtask 1 from Agent1.

[0375] Step 1509A: Agent2 understands the intent of subtask 1 and performs task planning.

[0376] Step 1510A: Agent2 invokes network capabilities from the NF network element.

[0377] For example, it can utilize wireless sensing capabilities to detect the surrounding environment.

[0378] Step 1511A: Agent2 performs dynamic path planning based on network capabilities.

[0379] Step 1512A: Based on the results of dynamic path planning, Agent2 translates the action semantics into underlying control commands and sends them to UE2.

[0380] Step 1513A: Agent1 sends the intent for subtask 2 to Agent3. Correspondingly, Agent3 receives the intent for subtask 2 from Agent1.

[0381] Step 1514A: Agent3 understands the intent of subtask 2 and performs task planning to determine the deployment of local intelligent agents to execute subtask 2.

[0382] Case 2) Identifier indicating the communication peer

[0383] Step 1504B: Agent1 understands the intent, decomposes the task based on the understood intent, and determines the agent capabilities required to complete the sub-tasks.

[0384] Step 1505B: Agent1 sends the identifier of the communication peer to the ARF network element. Correspondingly, the ARF network element receives the identifier of the communication peer from Agent1.

[0385] Step 1506B: The ARF network element determines the target intelligent agent implementation instance based on the identifier of the communication peer.

[0386] For example, if the identifiers of the communication peers are the identifier of UE2 and the identifier of UE3, then the ARF network element determines Agent2 corresponding to UE2 and Agent3 corresponding to UE3.

[0387] Step 1507B: The ARF network element sends the address of the target agent instance to Agent1. Correspondingly, Agent1 receives the address and capabilities of the agent instance from the ARF network element.

[0388] Step 1508B: Agent1 determines whether the target agent instance can execute the subtask based on the target agent instance's capabilities.

[0389] Taking Agent2 and Agent3 as target agent instances as an example, where Agent2 does not support local agents but Agent3 does, the method further includes the following, assuming the target agent instances are capable of executing subtasks:

[0390] Steps 1509B to 1515B are the same as steps 1508A to 1514A in the above embodiments, and will not be repeated here.

[0391] The following describes a possible embodiment of how, after the AI ​​agent function network element creates a corresponding intelligent agent instance for a user, subsequent terminals belonging to that user create corresponding intelligent agent instances, and how one of the multiple terminals triggers the target intelligent agent instance to perform tasks based on the capabilities of the target intelligent agent instance to achieve the intention.

[0392] As shown in Figure 16, Figure 16 is a schematic diagram of a terminal creating a corresponding intelligent agent instance according to an embodiment of this application. The first terminal is terminal 1, and the second terminals include terminal 2, terminal 3, and terminal 4. Terminal 1, terminal 2, and terminal 4 belong to user A, and terminal 3 belongs to user B.

[0393] Step 1: Terminal 1 sends a first registration request message to AIAF1. Correspondingly, AIAF1 receives the first registration request message from Terminal 1.

[0394] Step 2: AIAF1 creates the agent instance Agent1 corresponding to user A based on the capabilities of terminal 1.

[0395] In one possible implementation, when terminal 2 sends the first registration request message to AIAF1, AIAF1 determines to create a corresponding agent instance Agent1 for user A by determining that terminal 1 and terminal 2 belong to the same user.

[0396] In another possible implementation, AIAF1 can determine whether to create the corresponding agent instance Agent1 for user A by judging whether terminal 1 is the first terminal registered by user A.

[0397] Agent1 includes the agent instance Agent1a corresponding to terminal 1.

[0398] Step 3: Agent1 sends a second registration request message to the ARF network element. Correspondingly, the ARF network element receives the second registration request message from Agent1. The second registration request message includes Agent1's capabilities and identifier; Agent1's capabilities include the capabilities of Agent1a.

[0399] Step 4: Terminal 2 sends a first registration request message to AIAF1. Correspondingly, AIAF1 receives the first registration request message from Terminal 2. The first registration request message includes the capabilities of Terminal 2.

[0400] Step 5: AIAF1 creates an agent instance Agent1b corresponding to terminal 2 based on the capabilities of terminal 2 and associates it with Agent1.

[0401] Step 6: Agent1 sends a first registration update request message to the ARF network element. Correspondingly, the ARF network element receives the first registration update request message from Agent1. This first registration update request message is used to update the capabilities of Agent1, including the capabilities of Agent1b. The updated capabilities of Agent1 include the capabilities of Agent1a and Agent1b.

[0402] Step 7: Terminal 3 sends a first registration request message to AIAF2. Correspondingly, AIAF2 receives the first registration request message from Terminal 3. The first registration request message includes the local intelligent agent of Terminal 3.

[0403] Step 8: AIAF2 creates an agent instance Agent3 corresponding to terminal 3 based on the local agent of terminal 3.

[0404] Step 9: Agent3 sends a second registration request message to the ARF network element. Correspondingly, the ARF network element receives the second registration request message from Agent3. The second registration update request message includes Agent3's capabilities.

[0405] Step 10: Terminal 4 sends an access request message to the control plane of the core network. Correspondingly, the control plane receives the access request message from Terminal 4.

[0406] Step 11: The control plane sends a first registration request message to AIAF1 based on the access request message. Correspondingly, AIAF1 receives the first registration request message from the control plane. The first registration request message includes the capabilities of terminal 4.

[0407] Step 12: AIAF1 creates an agent instance Agent1c corresponding to terminal 4 based on the capabilities of terminal 4 and associates it with Agent1.

[0408] Step 13: Agent1 sends a second registration update request message to the ARF network element. Correspondingly, the ARF network element receives the second registration update request message from Agent1. This second registration update request message is used to update the capabilities of Agent1, including the capabilities of Agent1c. The updated capabilities of Agent1 include the capabilities of Agent1a, Agent1b, and Agent1c.

[0409] Based on the embodiment shown in Figure 16, the method provided in this application further includes: the intelligent agent instance corresponding to the first terminal executes a task through the target intelligent agent instance to achieve the intent initiated by the first terminal. The embodiment shown in Figure 17 differs from the embodiment shown in Figure 10 in that: when the target intelligent agent instance is the intelligent agent instance corresponding to the user, the intelligent agent instance corresponding to the user can directly allocate information for indicating subtasks to the target intelligent agent instance. The following is a specific embodiment, including:

[0410] Step 1: Terminal 1 sends an intent to Agent 1. Agent 1 then receives the intent from Terminal 1.

[0411] Step 2: Agent1 interprets the intent of Terminal 1 and breaks it down into multiple sub-tasks.

[0412] Step 3: Agent1 determines that Agent1b is capable of executing subtask 1 based on subtask 1, and performs instruction escaping based on subtask 1 to send low-level instructions to terminal 2.

[0413] Step 4: Agent1 determines the required capabilities based on subtask 2, and obtains the address of Agent3, which has the capability, by interacting with the ARF network element.

[0414] Step 5: Agent1 sends the intent for subtask 2 to Agent3.

[0415] Step 6: Agent3 invokes its capabilities based on the intent of subtask 2.

[0416] Step 7: Agent3 sends the intent of subtask 2 to terminal 3.

[0417] Step 8: Agent1 sends the intent of subtask 3 to Agent5 based on the address information of the third party, Agent5.

[0418] Step 9: Agent1 determines that Agent1c is capable of executing subtask 4 based on subtask 4, and performs instruction escaping based on subtask 4 to send low-level instructions to terminal 7.

[0419] In one possible implementation, as shown in Figure 18, the user's corresponding agent instance can use the core agent instance as the unified calling interface. The identifier of the core agent instance is the same as the identifier of the user's corresponding agent instance.

[0420] For example, the agent instance Agent1 corresponding to user A includes the core agent instance C-Agent. C-Agent can invoke Agent1a, Agent1b and Agent1c associated with Agent1 according to the intent.

[0421] The following describes the specific implementation method of the AI ​​agent function network element in creating an intelligent agent instance for subsequent terminals after creating the intelligent agent instance corresponding to the user:

[0422] As shown in Figure 19, Figure 19 is a schematic diagram illustrating a specific implementation method for a subsequent terminal to create a corresponding intelligent agent instance according to an embodiment of this application. The terminal supports a non-access layer. The specific method is as follows:

[0423] Steps 1901 to 1902 are the same as steps 1101 to 1102 in the above embodiments, and will not be repeated here.

[0424] Step 1903: UE2 sends the terminal's capabilities, terminal identifier, and digital identifier to the ISF network element. Correspondingly, the ISF network element receives the terminal capabilities, terminal identifier, and digital identifier from UE2. The digital identifier is used to indicate that UE2 is the user's subsequent terminal.

[0425] Step 1904: The ISF network element determines that UE2 is the user's terminal based on the digital identifier.

[0426] Step 1905: The ISF network element sends a request association message to the AIAF network element. Correspondingly, the AIAF network element receives the request association message from the ISF network element.

[0427] For example, the request association message includes UE2's capabilities, UE2's identifier, and a numerical identifier.

[0428] Optionally, if UE2 supports local agents, the request association message may also include local agent capabilities.

[0429] Steps 1906 to 1907 are the same as steps 1106 to 1107 in the above embodiments, and will not be repeated here.

[0430] Step 1908: The AIAF network element determines the intelligent agent template based on the terminal capabilities and subscription information, and creates an intelligent agent instance corresponding to UE2 based on the intelligent agent template and associates it with the intelligent agent instance corresponding to the user.

[0431] For example, AIAF creates Agent1b corresponding to UE2 and associates Agent1b with Agent1.

[0432] Step 1909: The AIAF network element sends the user's corresponding agent instance to the ISF network element. Correspondingly, the ARF network element receives the user's corresponding agent instance from the AIAF network element.

[0433] Step 1910: The ISF network element sends a first registration update request message to the ARF network element. Correspondingly, the ARF network element receives the first registration update request message from the ISF network element. The first registration update request message includes the updated capabilities and identifier of the agent instance corresponding to the user.

[0434] For example, the update registration message includes Agent1's capabilities and Agent1's identifier.

[0435] Step 1911: The ISF network element sends a registration response message to UE2. Correspondingly, UE2 receives the registration response message from the ISF network element.

[0436] The registration response message includes the capabilities of the updated agent instance.

[0437] Based on the specific implementation of updating the agent instance shown in Figure 19, and as shown in Figure 20, which is a schematic diagram of a specific implementation of a subsequent terminal registering and updating the agent instance according to an embodiment of this application, including:

[0438] Step 2001: UE2 interacts with ISF network elements to initiate the registration of intelligent services.

[0439] Step 2002: The ISF network element sends a deassociation request message to the AIAF network element. Correspondingly, the AIAF network element receives the deassociation request message from the ISF network element. The deassociation request message is used to request the deletion of the agent instance corresponding to UE2. The deassociation request message includes the agent instance's identifier and numerical identifier.

[0440] Step 2003: The AIAF network element deletes the agent instance corresponding to UE2 based on the deassociation request message and updates the agent instance corresponding to the user.

[0441] For example, the AIAF network element deletes Agent1b corresponding to UE2 based on the numerical ID2 and updates Agent1. In other words, Agent1 only associates with the intelligent agent instance Agent1a of UE1.

[0442] Step 2004: The AIAF network element sends an acknowledgment message to the ISF network element. Correspondingly, the ISF network element receives the acknowledgment message from the AIAF network element.

[0443] Step 2005: The AIAF network element sends a second registration update request message to the ARF network element. The ARF network element receives the second registration update request message from the AIAF network element. The second registration request message includes the identifier of the agent instance and the updated capabilities of the agent instance corresponding to the user.

[0444] Steps 2006 to 2008 are the same as steps 1206 to 1208 in the above embodiments, and will not be repeated here.

[0445] As shown in Figure 21, Figure 21 is a schematic diagram of another specific implementation method for creating an intelligent agent instance by a subsequent terminal (UE3) according to an embodiment of this application. The difference from the specific implementation method shown in Figure 17 is that the terminal does not support the non-access stratum. To distinguish the subsequent terminal, it is referred to as UE3 here, and the specific method is as follows:

[0446] Steps 2101 to 2104 are the same as steps 1301 to 1304 in the above embodiments, and will not be repeated here.

[0447] Step 2105: The SMF network element or UPF network element sends the terminal capabilities, terminal identifier, and digital identifier to the ISF network element. Correspondingly, the ISF network element receives the terminal capabilities, terminal identifier, and digital identifier from the SMF network element or UPF network element.

[0448] For example, the ISF network element's ability to receive UE3, the UE3's identifier, and the digital ID3.

[0449] Step 2106: The ISF network element determines UE3 as the user's terminal based on the digital identifier.

[0450] Step 2107: The ISF network element sends a request association message to the AIAF network element. Correspondingly, the AIAF network element receives the request association message from the ISF network element.

[0451] For example, the request association message includes UE3's capabilities, UE3's identifier, and the numeric ID3.

[0452] Optionally, if UE2 supports local agents, the request association message may also include local agent capabilities.

[0453] Steps 2108 to 2109 are the same as steps 1906 to 1907 in the above embodiments, and will not be repeated here.

[0454] Step 2110: The AIAF network element determines the intelligent agent template based on the terminal capabilities and subscription information, and creates an intelligent agent instance corresponding to UE3 based on the intelligent agent template and associates it with the intelligent agent instance corresponding to the user.

[0455] For example, AIAF creates Agent1c corresponding to UE3 and associates Agent1c with Agent1.

[0456] Step 2111: The AIAF network element sends a first registration update request message to the ARF network element. Correspondingly, the ARF network element receives the first registration update request message from the AIAF network element. The first registration update request message includes the updated capabilities and identifier of the agent instance corresponding to the user.

[0457] For example, the first registration update request message includes Agent1's capabilities and Agent1's identifier.

[0458] Step 2112: The ARF network element sends a registration response message to the ISF network element. Correspondingly, the ISF network element receives the registration response message from the AIAF network element.

[0459] The response message includes the updated capabilities of the agent instance corresponding to the user.

[0460] Based on the specific implementation of updating the agent instance shown in Figure 21, and as shown in Figure 22, which is a schematic diagram of another specific implementation of subsequent terminal registration of updating agent instance provided by the embodiments of this application, including:

[0461] Step 2201: UE3 interacts with network elements in the control plane and initiates PDU session termination.

[0462] Step 2202: The SMF network element sends a fifth request message to the ISF network element. Correspondingly, the ISF network element receives the fifth request message from the SMF network element. The fifth request message is used to request the cancellation of intelligent services.

[0463] The fifth request message includes the UE3 identifier and its numerical identifier.

[0464] Steps 2203 to 2207 are the same as steps 2002 to 2006 in the above embodiments, and will not be repeated here.

[0465] After a smart agent instance is created on a user's subsequent terminal, multiple smart agent instances, understanding the intent initiated by one terminal, execute the task. As shown in Figure 23, Figure 23 is a schematic diagram illustrating another specific implementation method of a smart agent instance corresponding to a first terminal executing a task through a target smart agent instance according to the terminal's intent, provided in an embodiment of this application. Taking terminal 1 (UE1) initiating an intent as an example, it specifically includes:

[0466] Steps 2301 to 2303 are the same as steps 1501 to 1503, and will not be repeated here.

[0467] Case 1) No identifier of the communication peer is indicated.

[0468] Step 2304A: Agent1 understands the intent, decomposes the task based on the understood intent, and determines that the capabilities of Agent1b associated with Agent1 meet the requirements for executing subtask 1. There is no agent instance that meets the requirements for executing subtask 2.

[0469] Step 2305A: Agent1 invokes network capabilities from the NF network element.

[0470] Step 2706A: Agent1 performs dynamic path planning based on network capabilities.

[0471] Step 2307A: Agent1 translates the action semantics into underlying control commands based on the results of dynamic path planning and sends them to UE2.

[0472] If the capabilities of the agent instance associated with Agent1 are insufficient to execute subtask 2, the following methods may be used:

[0473] Step 2308A: Agent1 sends the eighth request message to the ARF network element. Correspondingly, the ARF network element receives the eighth request message from Agent1.

[0474] The eighth request message requests the address of the target agent instance. The target agent instance is an agent instance that meets the required capabilities.

[0475] Step 2309A: The ARF network element determines the address of the target agent instance based on the eighth request message.

[0476] For example, the target intelligent agent instance is Agent3.

[0477] Step 2310A: The ARF network element sends the address of the target agent instance to Agent1. Correspondingly, Agent1 receives the address of the target agent instance from the ARF network element.

[0478] For example, the target intelligent agent instance is Agent3.

[0479] Steps 2311A to 2312A are the same as steps 1513A to 1514A in the above embodiments, and will not be repeated here.

[0480] Case 2) Identifier indicating the communication peer

[0481] The difference between step 2304B and step 2304A is that: Agent1 determines that the agent instance corresponding to the communication peer (UE2) is associated with Agent1 and meets the requirements for executing subtask 1. Alternatively, it determines that the agent instance corresponding to the communication peer (UE3) is associated with Agent1, but does not meet the requirements for executing subtask 2.

[0482] Steps 2305B to 2307B are the same as steps 2305A to 2307A, and will not be repeated here.

[0483] Step 2308B: Agent1 sends the identifier of the communication peer to the ARF network element.

[0484] For example, the communication peer is UE3. Although the agent instance corresponding to UE3 is associated with Agent1, it does not meet the requirement of executing subtask 2.

[0485] Steps 2309B to 2312B are the same as steps 2309A to 2312A, and will not be repeated here.

[0486] The above mainly describes the solutions of the embodiments of this application from the perspective of interaction between various network elements. It is understood that each network element, such as a terminal or network device, includes corresponding structures and / or software modules to perform the above functions in order to achieve them. Those skilled in the art should readily recognize that, based on the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed by hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0487] This application embodiment can divide functional units according to the terminal device and network device described above. For example, each function can be divided into separate functional units, or two or more functions can be integrated into one processing unit. The integrated unit can be implemented in hardware or as a software functional unit. It should be noted that the unit division in this application embodiment is illustrative and only represents one logical functional division; other division methods may be used in actual implementation.

[0488] The methods of the embodiments of this application have been described above with reference to Figures 5 and 6. The communication apparatus for executing the above methods provided in the embodiments of this application is described below. Those skilled in the art will understand that the methods and apparatus can be combined with and referenced in each other, and the communication apparatus provided in the embodiments of this application can execute the steps performed by the network device in the above analysis method.

[0489] When using an integrated unit, FIG24 shows the communication device involved in the above embodiment, which may include a communication module 2401 and a processing module 2402.

[0490] In an alternative implementation, the communication device 240 may further include a storage module 2403 for storing the program code and data of the communication device.

[0491] On one hand, the communication device 240 is an intelligent agent instance corresponding to the first terminal, or a chip applied to the intelligent agent instance corresponding to the first terminal. In this case, the communication module 2401 is used to support communication between the communication device and an external network element (e.g., the first terminal). For example, the communication module 2401 is used to perform signal transmission and reception operations of the intelligent agent instance corresponding to the first terminal in the above method embodiment. The processing module 2402 is used to perform signal processing operations of the intelligent agent instance corresponding to the first terminal in the above method embodiment.

[0492] In one example, the communication module 2401 is used to perform the receiving action executed by the intelligent agent instance corresponding to the first terminal in step 501 of FIG5 of the above embodiment.

[0493] In one example, the processing module 2402 is used to execute the processing actions performed by the intelligent agent instance corresponding to the first terminal in steps 502 to 503 of FIG5 of the above embodiment.

[0494] On the other hand, the communication device 240 is an artificial intelligence agent function network element, or a chip applied in an artificial intelligence agent function network element. In this case, the communication module 2401 is used to support communication between the communication device and external network elements (e.g., an agent registration function network element). For example, the communication module 2401 is used to perform signal transmission and reception operations of the artificial intelligence agent function network element in the above method embodiment. The processing module 2402 is used to perform signal processing operations of the artificial intelligence agent function network element in the above method embodiment.

[0495] In one example, the processing module 2402 is used to execute the processing actions performed by the artificial intelligence agent function network element in steps 601 to 602 of FIG6 of the above embodiment.

[0496] On the other hand, the communication device 240 is an agent registration function network element, or a chip applied in an agent registration function network element. In this case, the communication module 2401 is used to support communication between the communication device and external network elements (e.g., artificial intelligence agent function network elements). For example, the communication module 2401 is used to perform signal transmission and reception operations of the agent registration function network element in the above method embodiment. The processing module 2402 is used to perform signal processing operations of the agent registration function network element in the above method embodiment.

[0497] The processing module 2402 can be a processor or controller, such as a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field-programmable gate array, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. The processor can also be a combination that implements computational functions, such as a combination of one or more microprocessors, a combination of a digital signal processor and a microprocessor, etc. The communication module can be a transceiver, transceiver circuitry, or communication interface, etc. The storage module can be a memory.

[0498] When the processing module 2402 is a processor 2501 or a processor 2505, the communication module 2401 is a transceiver 2503, and the storage module 2403 is a memory 2502, the communication device involved in this application can be the communication device shown in FIG25.

[0499] Figure 25 shows a schematic diagram of the hardware structure of a communication device provided in an embodiment of this application. The hardware structure of the intelligent agent instance, the artificial intelligence agent function network element, and the intelligent agent registration function network element corresponding to the first terminal in this embodiment of the application can be referred to the structure shown in Figure 25. The communication device includes a processor 2501, a communication line 2504, and at least one transceiver (Figure 25 is only an example illustrating the inclusion of transceiver 2503).

[0500] The processor 2501 may be a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits used to control the execution of the program of the present application.

[0501] Communication line 2504 may include a path for transmitting information between the aforementioned components.

[0502] Transceiver 2503 uses any transceiver-like device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area network (WLAN), etc.

[0503] Optionally, the communication device may also include a memory 2502.

[0504] Memory 2502 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital versatile optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. Memory 2502 may exist independently and be connected to processor 2501 via communication line 2504. Memory 2502 may also be integrated with processor 2501.

[0505] The memory 2502 stores computer execution instructions for implementing the scheme of this application, and its execution is controlled by the processor 2501. The processor 2501 executes the computer execution instructions stored in the memory 2502, thereby implementing the communication method provided in the following embodiments of this application.

[0506] Optionally, the computer execution instructions in the embodiments of this application may also be referred to as application code, and the embodiments of this application do not specifically limit this.

[0507] In a specific implementation, as one example, processor 2501 may include one or more CPUs, such as CPU0 and CPU1 in FIG25.

[0508] In a specific implementation, as one example, the communication device may include multiple processors, such as processors 2501 and 2502 in Figure 25. Each of these processors may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. Here, "processor" may refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0509] Figure 26 is a schematic diagram of the structure of chip 260 provided in an embodiment of this application. Chip 260 includes one or more (including two) processors 2610 and communication interfaces 2630.

[0510] Optionally, the chip 260 also includes a memory 2640, which may include read-only memory and random access memory, and provides operation instructions and data to the processor 2610. A portion of the memory 2640 may also include non-volatile random access memory (NVRAM).

[0511] In some implementations, memory 2640 stores elements such as execution modules or data structures, or subsets thereof, or extended sets thereof.

[0512] In this embodiment of the application, the corresponding operation is executed by calling the operation instructions stored in the memory 2640 (the operation instructions can be stored in the operating system).

[0513] One possible implementation is that the intelligent agent instance corresponding to the first terminal, the artificial intelligence agent function network element, and the intelligent agent registration function network element have similar structures, and different devices can use different chips to implement their respective functions.

[0514] The processor 2610 controls the processing operations of any one of the intelligent agent instance, artificial intelligence agent function network element, and intelligent agent registration function network element corresponding to the first terminal. The processor 2610 can also be called a central processing unit (CPU).

[0515] Memory 2640 may include read-only memory and random access memory, and provides instructions and data to processor 2610. A portion of memory 2640 may also include NVRAM. For example, in an application, memory 2640, communication interface 2630, and memory 2640 are coupled together via bus system 2620, which may include, in addition to data bus, power bus, control bus, and status signal bus, etc. However, for clarity, all buses are labeled as bus system 2620 in Figure 26.

[0516] The methods disclosed in the embodiments of this application can be applied to or implemented by the processor 2610. The processor 2610 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware in the processor 2610 or by instructions in the form of software. The processor 2610 may be a general-purpose processor, a digital signal processor (DSP), an ASIC, a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules may be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 2640. Processor 2610 reads the information in memory 2640 and, in conjunction with its hardware, completes the steps of the above method.

[0517] The communication module described above can be a communication interface of the device, used to receive signals from other devices. For example, when the device is implemented as a chip, the communication module is the communication interface used by the chip to receive or send signals from other chips or devices.

[0518] On the one hand, a computer-readable storage medium is provided, which stores instructions that, when executed, implement the functions performed by the intelligent agent instance corresponding to the first terminal as shown in Figure 5.

[0519] On the one hand, a computer-readable storage medium is provided, in which instructions are stored, which, when executed, implement the functions performed by the artificial intelligence agent function network element as shown in Figure 6.

[0520] On the one hand, a computer-readable storage medium is provided, which stores instructions that, when executed, implement the functions performed by the intelligent agent registration function network element as shown in Figure 5.

[0521] On the one hand, a computer program product including instructions is provided. When the instructions are executed, they realize the function executed by the intelligent agent instance corresponding to the first terminal as shown in Figure 5.

[0522] On the one hand, a computer program product including instructions is provided. When the instructions are executed, the functions executed by the artificial intelligence agent function network element as shown in Figure 6 are realized.

[0523] On the one hand, a computer program product including instructions is provided. When the instructions are executed, the functions performed by the intelligent agent registration function network element as shown in Figure 5 are realized.

[0524] On the one hand, a chip is provided, which is applied in the intelligent agent instance corresponding to the first terminal. The chip includes at least one processor and a communication interface. The communication interface and at least one processor are coupled. The processor is used to run instructions to realize the functions executed by the intelligent agent instance corresponding to the first terminal as shown in Figure 5.

[0525] On the one hand, a chip is provided, which is applied in the intelligent agent instance corresponding to the first terminal. The chip includes at least one processor and a communication interface. The communication interface and at least one processor are coupled. The processor is used to run instructions to realize the functions executed by the artificial intelligence agent function network element as shown in Figure 6.

[0526] On the one hand, a chip is provided, which is applied in the intelligent agent instance corresponding to the first terminal. The chip includes at least one processor and a communication interface. The communication interface and at least one processor are coupled. The processor is used to run instructions to realize the functions executed by the intelligent agent registration function network element as shown in Figure 5.

[0527] This application provides a communication system comprising: an artificial intelligence agent function network element and an intelligent agent registration function network element. The artificial intelligence agent function network element includes an intelligent agent instance corresponding to a first terminal.

[0528] The agent instance corresponding to the first terminal is used to execute the functions shown in Figure 5. The AI ​​agent function network element is used to execute the functions shown in Figure 6. The agent registration function network element is used to execute the functions shown in Figure 5.

[0529] The explanations and beneficial effects of the relevant content in any of the communication devices provided above can be found in the corresponding method embodiments provided above, and will not be repeated here.

[0530] In this embodiment, the intelligent agent instance, artificial intelligence agent function network element, or intelligent agent registration function network element corresponding to the first terminal includes a hardware layer, an operating system layer running on top of the hardware layer, and an application layer running on top of the operating system layer. The hardware layer includes hardware such as a central processing unit (CPU), a memory management unit (MMU), and memory (also called main memory). The operating system can be any one or more computer operating systems that implement business processing through processes, such as Linux, Unix, Android, iOS, or Windows. The application layer includes applications such as browsers, address books, word processing software, and instant messaging software. Furthermore, the embodiments of this application do not particularly limit the specific structure of the execution subject of the method provided in the embodiments of this application. As long as it is possible to communicate according to the method provided in the embodiments of this application by running a program that records the code of the method provided in the embodiments of this application, for example, the execution subject of the method provided in the embodiments of this application may be the intelligent agent instance corresponding to the first terminal, or the artificial intelligence agent function network element, or the intelligent agent registration function network element, or it may be a functional module in the intelligent agent instance, or the artificial intelligence agent function network element, or the intelligent agent registration function network element corresponding to the first terminal that can call and execute the program.

[0531] Furthermore, various aspects or features of this application can be implemented as methods, apparatus, or articles of manufacture using standard programming and / or engineering techniques. The term "article of manufacture" as used herein encompasses a computer program accessible from any computer-readable device, carrier, or medium. For example, computer-readable media may include, but are not limited to: magnetic storage devices (e.g., hard disks, floppy disks, or magnetic tapes), optical discs (e.g., compact discs (CDs), digital versatile discs (DVDs), etc.), smart cards, and flash memory devices (e.g., erasable programmable read-only memory (EPROMs), cards, sticks, or key drives, etc.). Additionally, the various storage media described herein may represent one or more devices and / or other machine-readable media for storing information. The term "machine-readable medium" may include, but is not limited to, wireless channels and various other media capable of storing, containing, and / or carrying instructions and / or data.

[0532] It should be understood that the processor mentioned in the embodiments of this application can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0533] It should also be understood that the memory mentioned in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchlink DRAM (SLDRAM), and Direct Rambus RAM (DR RAM).

[0534] It should be noted that when the processor is a general-purpose processor, DSP, ASIC, FPGA, or other programmable logic device, discrete gate or transistor logic device, or discrete hardware component, the memory (storage module) is integrated into the processor.

[0535] It should be noted that the memories described herein are intended to include, but are not limited to, these and any other suitable types of memories.

[0536] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0537] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0538] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0539] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0540] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0541] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0542] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A communication method of an agent network, characterized by, The method comprises: receiving indication information from a first terminal, the indication information being used to indicate an intention of a user, the user being associated with the first terminal; determining a target agent instance according to the indication information, the capability of the target agent instance comprising a capability of a terminal corresponding to the target agent instance and a capability given to the terminal by a network; triggering the target agent instance to perform a task according to the capability of the target agent instance to achieve the intention of the user.

2. The method of claim 1, wherein, The determining of the target agent instance according to the indication information comprises: determining the task according to the intention of the user; the determining of the target agent instance comprises determining an agent instance having a capability for performing the task.

3. The method of claim 2, wherein, The task comprises a first sub-task, and the determining of the target agent instance comprises determining an agent instance having a capability for performing the task, which comprises: determining the capability for performing the first sub-task; determining that the target agent instance comprises a first agent instance having the capability for performing the first sub-task.

4. The method of claim 3, wherein, The determining of the target agent instance comprising a first agent instance comprises: sending a request message to an agent registration function network element, the request message being used to indicate the capability for performing the first sub-task; receiving information of the first agent instance from the agent registration function network element.

5. The method of claim 3, wherein, The determining of the target agent instance comprising a first agent instance comprises: determining the first agent instance from at least one agent instance associated with an agent instance corresponding to the first terminal.

6. The method of claim 5, wherein, The at least one agent instance associated with the agent instance corresponding to the first terminal comprises that terminals corresponding to the at least one agent instance belong to the user.

7. The method according to any one of claims 1 to 6, characterized in that, The indication information comprises an identifier of a second terminal, and the determining of the target agent instance according to the indication information comprises: determining that the target agent instance comprises an agent instance corresponding to the second terminal according to the identifier of the second terminal.

8. The method according to any one of claims 3-6, characterized in that, The triggering of the target agent instance to perform a task according to the capability of the target agent instance to achieve the intention comprises: sending information describing the first sub-task to the target agent instance.

9. The method of claim 8, wherein, The method further comprises: receiving the information describing the first sub-task; in response to the information describing the first sub-task, invoking the capability of the target agent instance to perform the first sub-task.

10. The method of claim 9, wherein, The invoking of the capability of the target agent instance to perform the first sub-task comprises: planning and combining the capability of the terminal according to the capability given to the terminal by the network to perform the first sub-task.

11. A communication method of an agent network, characterized by, The method comprises: creating an agent instance corresponding to a first terminal; indicating the agent instance to send a registration request message to an agent registration function network element, the registration request message comprising a capability of the agent instance, the capability of the agent instance comprising a capability of the first terminal and a capability given to the first terminal by a network.

12. The method of claim 11, wherein, Before the creating of the agent instance corresponding to the first terminal, the method further comprises: receiving information from the first terminal, the information being used to indicate a capability of the first terminal; the creating the agent instance corresponding to the first terminal comprises: creating the agent instance corresponding to the first terminal according to the capability of the first terminal.

13. The method of claim 12, wherein, the creating the agent instance corresponding to the first terminal according to the capability of the first terminal comprises: determining an agent template according to the capability of the first terminal and subscription information of the first terminal; creating the agent instance corresponding to the first terminal according to the agent template.

14. The method according to any one of claims 11 to 13, characterized in that, the method further comprises: receiving information from a second terminal, the information being used to indicate a capability of the second terminal; in a case where the first terminal and the second terminal belong to a same user, creating an agent instance corresponding to the user, a capability of the agent instance corresponding to the user comprising a capability of the agent instance corresponding to the first terminal and a capability of the agent instance corresponding to the second terminal.

15. The method of claim 14, wherein, the method further comprises: indicating the agent instance corresponding to the user to send a second request message to the network element of the agent registration function, the second request message being used to indicate the capability of the agent instance corresponding to the user.

16. A method of communication for a network of intelligent agents, the method comprising: the method comprises: receiving a request message from an agent instance corresponding to a first terminal, the request message being used to indicate a capability for performing a task, the task belonging to a task needed to be performed to achieve an intent of a user; sending information of a first agent instance to the agent instance corresponding to the first terminal, a capability of the first agent instance comprising a capability of a terminal corresponding to the first agent instance and a capability of the network for the terminal, the capability of the first agent instance comprising the capability for performing the task.

17. The method of claim 16, wherein, the method comprises: receiving a registration request message from the agent instance corresponding to the first terminal, the registration request message comprising a capability of the agent instance.

18. The method of claim 17, wherein, the method further comprises: receiving a second request message from the agent instance corresponding to the user, the second request message being used to indicate a capability of the agent instance corresponding to the user, the capability of the agent instance corresponding to the user comprising a capability of the agent instance corresponding to the first terminal and a capability of the agent instance corresponding to the second terminal.

19. A communications device, characterized by The apparatus comprises modules for performing the method of any of claims 1-18.

20. A communications device, characterized by The communication device comprises a memory and a processor, the memory is used to store instructions, the processor is used to execute the instructions stored in the memory, and execution of the instructions stored in the memory causes the processor to perform the method of any of claims 1-10, or perform the method of any of claims 11-15, or perform the method of any of claims 16-18.

21. A chip, characterized by The chip comprises at least one processor and a communication interface, the communication interface is coupled with the at least one processor, the at least one processor is used to run computer programs or instructions to realize the method as claimed in any one of claims 1-10, or realize the method as claimed in any one of claims 11-15, or realize the method as claimed in any one of claims 16-18, and the communication interface is used to communicate with other modules outside the chip.

22. A computer-readable storage medium, characterized in that, The computer readable storage medium stores instructions, when the instructions are run, the method as claimed in any one of claims 1-10 is realized, or the method as claimed in any one of claims 11-15 is realized, or the method as claimed in any one of claims 16-18 is realized.

23. A computer program product comprising instructions, characterized in that, When the instructions are run on the computer, the computer is caused to execute the method as claimed in any one of claims 1-10, or execute the method as claimed in any one of claims 11-15, or execute the method as claimed in any one of claims 16-18.

24. A communication system, characterized by The system comprises at least one of the following: a device for executing the method as claimed in any one of claims 1-10; a device for executing the method as claimed in any one of claims 11-15; or a device for executing the method as claimed in any one of claims 16-18.

Images