Intelligent agent-driven 5G customized network opening method and system, electronic equipment and storage medium

Abstract

The invention provides an agent-driven 5G customized network opening method and system, electronic equipment and a storage medium, and relates to the field of automatic network opening. The method comprises the following steps: identifying a demand and a business scene through multiple rounds of conversations between a central control agent and a user, calling a parameter input agent to guide parameter input, verifying supplementary parameters through a data processing agent, generating an opening scheme by a scheme generation agent, and calling a network opening service by a service orchestration agent to complete opening. The content arrangement agent feeds back a result; the system comprises a central control agent, a unitized agent set, a network opening scheme and service model module and a capability pool. According to the invention, automatic opening of the 5G customized network is realized, multi-scene complex services are adapted, and the service support period is shortened.

Description

Technical Field

[0001] This invention relates to the field of automated network activation, and more particularly to a method, system, electronic device, and storage medium for 5G customized network activation driven by an intelligent agent. Background Technology

[0002] As a relatively new technology or trend, the development of large-scale intelligent agents is currently mainly focused on expanding and optimizing the various capabilities of the agents themselves, as well as applications in some public scenarios. Applications are still relatively limited in vertical sectors and smaller-scale business scenarios. Compared to the general 5G networks used by the public, customized 5G networks tailored to specific enterprise needs have diverse requirements for network customization. To meet the different needs of customers, in practice, network element parameter configuration is still done through a combination of system and manual operation and maintenance. In the past two years, with the continuous development of business, customized 5G networks have gradually accumulated some commonly used general scenarios. Although the details of network element configuration parameters may differ, the configuration types are universal and have the conditions for automated processing.

[0003] Currently, there are no applications specifically for network activation based on large-scale intelligent agents. For simple network activation, most parameters are fixed, with only a dozen or so parameters changing, and this can be completed through a user interface. For complex network activation, due to the large number of parameters and commands, most of these tools are auxiliary tools for network element maintenance personnel. Summary of the Invention

[0004] Purpose of the invention: To propose a method, system, electronic device and storage medium for enabling 5G customized networks driven by intelligent agents, so as to solve the above-mentioned problems existing in the prior art.

[0005] In a first aspect, the present invention proposes a method for enabling a customized 5G network driven by an intelligent agent, comprising the following steps: The central control intelligent agent engages in multiple rounds of dialogue with users to identify their network needs and the corresponding 5G customized network service scenarios. The central control intelligent agent calls the parameter input intelligent agent, which inputs the configuration according to the preset business scenario requirements, obtains the input order and format of the required parameters, and guides the user to input network activation parameters through dialogue; for drop-down type parameters, it calls the remote plugin to obtain the data dictionary; for parameters related to previous and subsequent businesses, it calls the remote plugin to obtain the subsequent parameter rules according to the input information. The central control intelligent agent calls the data processing intelligent agent to verify the validity of the parameters obtained in step S2; if the verification passes, the data processing intelligent agent supplements the data integrity according to the network opening service input parameter requirements, including calling the existing network resource survey capability to query the existing network data; if the verification fails, it returns to the central control intelligent agent to guide the user to re-enter the parameters. The central control intelligent agent calls the solution generation intelligent agent. The solution generation intelligent agent is based on the solution configuration and solution template in the preset network activation solution model. It fills the parameters after verification and supplementation in step S3 into the template, and calls the large model language generation capability to generate at least one 5G customized network activation solution for user confirmation. After the user confirms the plan, the central control intelligent agent calls the service orchestration intelligent agent. The service orchestration intelligent agent calls the corresponding network activation service and network element operation commands according to the deterministic capability configuration and acquisition and control capability orchestration in the preset network activation service model to complete the 5G customized network activation. The central control agent calls the content orchestration agent, which, based on the prompts and semantic constraints configured in the knowledge base, invokes the large model's logical dialogue capabilities to organize the activation result into a user-understandable script and provides feedback to the user.

[0006] In a further embodiment of the first aspect, the solution generating agent generates at least two network activation schemes for users to choose from; the solution configuration has at least two scheme templates pre-set for each service scenario, including a high-load rate guarantee template and a mobility rate guarantee template.

[0007] In a further embodiment of the first aspect, the parameter input agent, data processing agent, content orchestration agent, solution generation agent, and service orchestration agent are designed as modular units that can be called individually or in combination.

[0008] In a further embodiment of the first aspect, the network activation scheme model forms the smallest scheme unit by extracting upper-layer application scenarios from top to bottom, and the network activation service model forms the 5G customized network activation service unit by summarizing network element capabilities from bottom to top.

[0009] A second aspect of the present invention provides a 5G customized network activation system, the system comprising: The central control intelligent agent is used to engage in multi-turn dialogues with users to identify network demand intentions and business scenarios, and to orchestrate and schedule various unit-based intelligent agents and external capabilities. The unitized intelligent agent set includes parameter input intelligent agent, data processing intelligent agent, content arrangement intelligent agent, solution generation intelligent agent and service arrangement intelligent agent, which respectively perform parameter input, data verification and supplementation, result script organization, activation solution generation and network activation service invocation; The network activation solution model module stores the configuration of requirement input, configuration of existing network resource survey, and solution configuration. The solution configuration includes the relationship between business scenarios and solution templates. The network activation service model module stores deterministic capability configuration and acquisition and control capability orchestration, wherein the acquisition and control capability orchestration includes network element operation commands, execution order and required parameters; The capability pool includes a knowledge base, remote plugins, acquisition and control installation interfaces, acquisition and control disassembly interfaces, DCOOS platform APIs, and resource sharing query interfaces, providing data support and interface call capabilities for each intelligent agent.

[0010] In a further embodiment of the second aspect, the parameter input agent includes a requirement input configuration unit for storing the parameter input order, format, and dialogue skills corresponding to each business scenario. The parameter input agent guides the user to input parameters through this unit.

[0011] In a further embodiment of the second aspect, the data processing agent includes a legality verification unit and a data supplementation unit; the legality verification unit verifies the legality of the parameters, and the data supplementation unit calls the existing network resource survey interface to obtain existing network data to supplement the completeness of the parameters.

[0012] In a further embodiment of the second aspect, the solution generation agent includes a solution configuration unit and a solution template library; the solution template library stores high-load rate guarantee templates, mobility rate guarantee templates, and ULCL enhanced templates.

[0013] In a further embodiment of the second aspect, the service orchestration agent includes a deterministic capability configuration unit and a data acquisition and control capability invocation unit; the deterministic capability configuration unit stores the service invocation order and required parameters corresponding to the network activation scheme, and the data acquisition and control capability invocation unit invokes the data acquisition and control interface to perform network element operations.

[0014] A third aspect of the present invention provides an electronic device comprising: a processor and a memory storing computer program instructions; wherein the processor, when executing the computer program instructions, implements the agent-driven 5G customized network activation method described in the first aspect.

[0015] In a fourth aspect, the present invention provides a computer-readable storage medium storing at least one executable instruction that, when executed on an electronic device, causes the electronic device to perform the agent-driven 5G customized network activation method described in the first aspect.

[0016] Beneficial effects: (1) This invention uses large model intelligent agent related technologies. It not only solves the existing business problems, but also divides the intelligent agent into units in combination with the business. These intelligent agent units can be used in combination or individually, so that they can quickly build new capabilities in complex applications with multiple scenarios and multiple parameters, quickly adapt to new business scenarios, and greatly shorten the business support cycle and reduce labor costs.

[0017] (2) This method extracts the scenarios of upper-layer applications from top to bottom, forming the smallest solution unit, which can quickly assemble and support similar scenarios; it summarizes the lower-layer network from bottom to top, forming the network activation service unit of 5G customized network, laying the foundation for the automated network activation of 5G customized network, accelerating the network activation time, and saving the labor cost of network activation. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the 5G customized network activation process in the embodiment.

[0019] Figure 2 This is an architecture diagram of the 5G customized network activation system in the embodiment.

[0020] Figure 3 This is a schematic diagram of the network activation scheme model and network activation service model in the embodiment. Detailed Implementation

[0021] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the invention. However, it will be apparent to those skilled in the art that the invention can be practiced without one or more of these details. In other instances, certain technical features well-known in the art have not been described in order to avoid obscuring the invention.

[0022] This invention discloses a method and system for 5G customized network activation driven by a large model intelligent agent. It mainly uses a multi-turn dialogue of a large model intelligent agent to confirm the customer's network scenario requirements, guide the customer to confirm some network parameters according to the scenario, provide the customer with a corresponding network activation solution, and assist the customer in activating the network according to the solution after the customer adjusts and confirms the solution, thus completing the automated network activation.

[0023] Before describing the embodiments, some technical terms will be explained: Large-scale models: These typically refer to machine learning models with extremely large scale and complexity. Composed of billions or even more parameters, and trained using deep learning and massive datasets, large-scale models can possess a degree of dialogue and logical reasoning ability, and can even interact based on the context of the chat, truly engaging in conversation like humans. Since the advent of ChatGPT in late 2022, large-scale models have gradually permeated all aspects of society, changing the way people live and work. Intelligent agent: It is an intelligent processing unit based on a large model, with powerful language communication and logical reasoning capabilities. It can automatically complete some complex tasks. In the process of completing tasks, it can autonomously analyze data, perform logical reasoning, and perform remote invocation and other operations.

[0024] Network activation: This mainly refers to the activation of networks for 5G2B services, i.e., the activation of customized 5G networks. These are typically tailored to the specific network needs of government and enterprise customers. The process mainly involves communicating the enterprise's requirements, providing the customer with several network activation solutions, and after confirmation with the customer, the operations engineer activates the network elements. These typically include UDM, PCF (used to configure billing policies), SMF, UPF, and may also involve DCGW (service switch).

[0025] This patent belongs to the field of large-scale intelligent agent applications and automated network activation.

[0026] Large model intelligent agent application: A large model is an intelligent algorithm application that can have dialogue and logical reasoning capabilities. With the help of large models, intelligent agents can automatically complete some complex tasks. As a relatively new technology, large models are gradually empowering various business fields.

[0027] Automated network commissioning: For 5G2B services, namely 5G customized networks for enterprises, network commissioning varies due to different enterprise needs. Currently, maintenance is mostly done manually. As the business volume increases, we can gradually build some business models for some common scenarios, incorporate the different needs of users, and complete differentiated customization. During this process, communicating with customers about their needs and developing modification plans has become the main task.

[0028] Leveraging the large model's ability to converse and reason, it perfectly complements the automated network activation process, enabling fully online automated network activation through multi-turn dialogues.

[0029] This patent primarily addresses the issue of rapidly activating networks for complex network needs of government and enterprise clients. The main steps are: using a large-scale intelligent model to identify the user's intent and network requirement parameters through multi-turn dialogue; using the large-scale intelligent model to rationally verify and assemble the parameters; and finally, using the large-scale intelligent model to orchestrate and invoke the actual network activation service and return the relevant information to the user.

[0030] The main invention lies in: 1. Based on large-scale model and agent-related technologies, network activation capabilities were achieved through multi-round dialogue with users. The agents were generalized and modularized, with a central control agent orchestrating and scheduling the overall process, calling upon parameter input agents, content orchestration agents, data processing agents, solution generation agents, and service orchestration agents to complete application services. This agent capability not only provides network activation for customized 5G networks but also rapidly supports complex services with similar parameters across multiple scenarios.

[0031] 2. For common business scenarios, from a business perspective, establish a template model capability for network activation schemes based on 5G customized networks. From a network element perspective, referencing the acquisition and control API, establish a network activation orchestration unit based on 5G customized network services and a standard capability pool based on network elements. This provides basic model support for calling large models. This model capability support not only provides capability support for large models but also provides atomic capability support for other similar scenarios.

[0032] Example 1 like Figure 1 As shown, a method for enabling a 5G customized network driven by a large-scale intelligent agent includes the following steps: S1. Demand Recognition: The central control intelligent agent engages in multi-round dialogues with the user through a natural language interaction interface. Based on the intent recognition capability of the large model, it parses the text information input by the user, extracts network demand keywords (such as bandwidth, latency, and number of connections), and matches them with preset business scenario tags (such as industrial control scenario, smart healthcare scenario, and vehicle networking scenario) to complete the recognition of the user's network demand intent and the corresponding business scenario.

[0033] S2. Parameter Input: The central control agent calls the parameter input agent based on the identified business scenario. The parameter input agent reads the corresponding business scenario requirement input configuration stored in the network activation scheme model, and obtains the input order of the required parameters in this scenario (e.g., inputting regional information first, then inputting bandwidth requirements), data format (e.g., IP address uses IPv4 format, bandwidth unit is Mbps) and dialogue guidance strategy.

[0034] During parameter input, for drop-down parameters (such as operator selection and frequency band selection), the parameter input agent calls remote plugins in the capability pool, obtains the latest data dictionary through the resource sharing query interface, and displays it to the user for selection; for related parameters (such as access point and available bandwidth), based on the information already entered by the user, the remote plugin is called to query the corresponding parameter constraint rules, and the input range of subsequent parameters is dynamically adjusted (e.g., when the user selects a specific access point, only the bandwidth options supported by that access point are displayed).

[0035] S3. Data Processing: The central control intelligent agent sends the set of parameters entered by the user to the data processing intelligent agent. The data processing intelligent agent first performs a validity check, including: Format verification: Checks whether parameters such as IP address and MAC address conform to the standard format; Range verification: Verify whether parameters such as bandwidth and latency are within the range allowed by the technical specifications; Logical verification: Confirm whether the correlation between parameters is reasonable (e.g., whether the selected frequency band and bandwidth match).

[0036] If the verification fails, the data processing agent generates an error message and returns it to the central control agent, which then guides the user to re-enter the parameters. If the verification passes, the data processing agent calls the live network resource survey interface in the capability pool to query live network data (such as base station coverage and idle channel resources in the target area) according to the network activation service input parameter requirements, and supplements and improves the parameter set (such as automatically filling in base station ID and channel number).

[0037] S4. Solution Generation: The central control agent invokes the solution generation agent, which reads the solution configuration from the network commissioning solution model and obtains the set of solution templates corresponding to the current business scenario (e.g., low-latency templates and high-reliability templates for industrial control scenarios). The complete set of parameters after data processing is filled into each solution template according to the field mapping relationship. The language generation capability of the large model is then used to perform natural language conversion and optimization on the template content, generating at least one 5G customized network commissioning solution that includes network architecture, resource configuration, and deployment steps. This solution is then displayed to the user through the central control agent.

[0038] S5. Service Orchestration: After the user confirms the activation plan, the central control agent sends a plan confirmation command to the service orchestration agent. The service orchestration agent reads the deterministic capability configuration in the network activation service model, parses the service call sequence corresponding to the plan, and determines the execution order of network element operation commands (e.g., configure UDM first, then PCF, and finally UPF) and parameter passing rules based on the acquisition and control capability orchestration.

[0039] The service orchestration agent sequentially calls the corresponding network activation services and network element operation commands (such as the user subscription data configuration command of UDM, the policy control command of PCF, and the session routing configuration command of UPF) through the acquisition and control interface (acquisition and control installation interface) in the capability pool, and obtains the execution results of each step in real time and performs status monitoring.

[0040] S6. Result Feedback: After service orchestration is completed, the central control agent calls the content orchestration agent. The content orchestration agent reads the result feedback prompt templates and semantic constraints (such as avoiding technical terms and using structured expressions) stored in the knowledge base, and uses the logical dialogue capabilities of the large model to convert the activation results (including success / failure status, configuration details, and network test data) into natural language statements that the user can understand, and then feeds them back to the user through the central control agent's interactive interface.

[0041] Example 2 like Figure 2As shown, a 5G customized network commissioning system driven by a large-scale model intelligent agent, implementing the above method, is disclosed. The overall system architecture is divided into three layers. The bottom layer is the foundation capability, providing the upper layers with basic AI capabilities for large and small models, data capabilities related to customized networks, and basic technical capabilities. The middle layer is the business capability layer, which mainly processes the underlying basic capabilities to form business capabilities, which are then orchestrated and organized by intelligent agents to provide business support capabilities to the upper layers. The top layer is the business scenario layer, which supports actual business by calling business capabilities.

[0042] The central control intelligent agent adopts a modular architecture design, comprising a dialogue management unit, a scene recognition unit, and an intelligent agent scheduling unit. The dialogue management unit is responsible for maintaining the multi-turn dialogue context with the user; the scene recognition unit integrates a large model intent recognition interface to match user needs with business scenarios; and the intelligent agent scheduling unit sends invocation instructions to each modular intelligent agent and receives the return results based on preset business process rules.

[0043] Unitary intelligent agent set: The parameter input agent comprises a requirement input configuration unit and a dialogue guidance unit. The requirement input configuration unit stores the parameter attributes (name, type, required fields) and input rules corresponding to each business scenario; the dialogue guidance unit generates natural language guidance statements based on the parameter attributes, receives user input, and performs format conversion.

[0044] The data processing intelligent agent consists of a validity verification unit and a data supplementation unit. The validity verification unit has a built-in parameter verification rule library and performs format, range, and logic verification; the data supplementation unit obtains and fills in missing network resource parameters by calling the existing network resource survey interface.

[0045] Solution generation intelligent agent: includes solution configuration unit and solution template library. Solution configuration unit stores the mapping relationship between business scenarios and solution templates; solution template library contains standardized solution frameworks, structured according to network functional modules (such as core network, access network, and transmission network).

[0046] The service orchestration agent consists of a deterministic capability configuration unit and a data acquisition and control capability invocation unit. The deterministic capability configuration unit stores the correspondence between activation schemes and service invocation sequences; the data acquisition and control capability invocation unit sends network element operation commands through the DCOOS platform API and receives the execution results.

[0047] Content orchestration agent: consists of a script template unit and a semantic optimization unit. The script template unit stores script frameworks for different result types (success, failure, warning); the semantic optimization unit calls the large model interface to perform natural language conversion on the original results.

[0048] Network activation scheme model module: Stored using a structured database, including: Requirement entry configuration table: records business scenario ID, parameter ID, entry order, and data format; Live network resource survey configuration table: stores survey interface addresses, parameter mapping relationships, and update frequency; Solution Configuration Table: Associates business scenario ID with solution template ID, and records template priority.

[0049] Network activation service model module: adopts a distributed storage architecture, including: Deterministic capability configuration library: stores storage service ID, execution conditions, input / output parameters; Acquisition and control capability orchestration library: records network element type (UDM, PCF, SMF, UPF, DCGW), operation command ID, execution order, and dependencies.

[0050] Capability Pool: Integrates various interfaces and resources, including: Knowledge base: Stores business scenario descriptions, parameter specifications, script templates, and error code lookup tables; Remote plugins include a data dictionary query plugin, a parameter rule parsing plugin, and a resource survey plugin. Acquisition and control interface: Provides acquisition and control installation interface (for performing network element configuration) and acquisition and control disassembly interface (for abnormal rollback); Open APIs: including DCOOS platform API (for network element management) and resource sharing query API (for resource data acquisition).

[0051] Baseline Capability Layer: Provides fundamental support for the system, including: AI Foundation: Deploy large model services (providing natural language processing and logical reasoning capabilities) and small model services (providing parameter validation and rule matching capabilities); Data foundation: includes a real-time database (stores dialogue data and execution status), a summary database (stores historical activation records), and a derivative database (stores statistical analysis results); Technical foundation: Provides Remote Procedure Call (RPC) service, rule engine service (for executing business rule matching) and document operation service (for processing solution templates and script templates).

[0052] Example 3 This invention mainly comprises three functional modules: intelligent acquisition of intent parameters and capacity scheduling, establishment and management of network activation scheme models, and establishment and management of network activation service models.

[0053] Intelligent acquisition of intent parameters and capability scheduling, such as Figure 1As shown, it mainly utilizes the capabilities of large-scale intelligent agents to obtain user intent and corresponding parameters through multi-turn dialogues, and completes the activation of relevant 5G customized networks; the central control intelligent agent controls the entire process, calling the capabilities of unitized intelligent agents and external capability pools to complete relevant capability operations. Figure 1 middle: Central Control Agent: The central control agent is a control unit that completes network activation through multi-turn dialogue. Its main functions include identifying the relevant scenario through multi-turn dialogue and, based on actual needs, calling corresponding parameter input agents, data processing agents, content orchestration agents, solution generation agents, service orchestration agents, and related external capabilities. Parameter input agent: Based on the business scenario, it retrieves the required parameters and their order from the requirement input configuration and guides the user to input parameters through dialogue. Simple parameters can be entered directly; drop-down parameters will call the corresponding remote plugin to retrieve the data dictionary. For parameters related to previous and subsequent business operations, it will call the corresponding remote plugin to retrieve the subsequent parameter rules based on the already entered information. The content orchestration agent: It takes the acquired parameters, utilizes the logical analysis capabilities of the large model, and organizes them according to the format of subsequent input parameters; it takes the returned data, according to the relevant prompts configured in the knowledge base and the semantic constraints on the returned statements, utilizes the relevant logical dialogue capabilities of the large model, and organizes the results into language that is easy for the user to understand. Data processing intelligent agent: Based on the business scenario, it performs legality verification on the input parameters and supplements the data with completeness according to the requirements of the capability input parameters. For example, based on the business scenario, it calls the existing network resource survey capability to supplement the data. For those that fail the verification, it will return to the central control intelligent agent to continue entering parameters.

[0054] Solution generation agent: Based on the verified parameters, according to the solution template configured in the solution capability configuration and the semantic constraints of the solution generation script, the agent fills the relevant parameters into the template, calls the large model to generate several corresponding network activation solutions through relevant language generation capabilities, and provides customers with a choice.

[0055] Service orchestration agent: Based on the network activation plan selected by the user, the central control agent calls the corresponding network activation service capabilities to perform the actual network activation.

[0056] The network activation scheme model and the network activation service model are respectively top-down models from the user's perspective to the network activation scheme, and bottom-up models from the network element to network activation. See Figure 3 .

[0057] The network activation scheme model mainly includes: 1. Configuration of input parameters: This configuration allows the large model agent to obtain the corresponding parameter input order and format, as well as dialogue techniques, and then obtain parameters through multi-turn dialogue. 2. Live Network Resource Survey and Configuration: This configuration allows the large-scale model agent to query relevant live network data via remote plugins, providing data support for solution generation. 3. Solution Configuration: The configuration sets the relationship between scenarios and solutions, specifying several related solutions available for each scenario and providing templates. After obtaining parameters and live network data, the corresponding solution template for each scenario is invoked to generate the solution. The network activation service model mainly includes: 1. Acquisition and Control Capability Orchestration: This refers to the network elements involved in the network activation service, as well as the corresponding operation commands, the order of command execution, and the required parameters. This capability will directly generate the corresponding commands for the network elements and send them to the acquisition and control system in sequence to complete the operation of the corresponding network element capabilities.

[0058] 2. Deterministic capability configuration: This refers to the basic network activation service capabilities that need to be invoked according to the network activation plan after the user has determined the plan, the activation order, and the required parameters.

[0059] Application Examples Taking the launch of a customized 5G network in an industrial park as an example, the implementation process of this invention is as follows: Users input "Deploy a 5G network for the park, supporting simultaneous connections of 200 devices with latency controlled within 10ms" through the interactive interface. The central control intelligent agent confirms the user's needs through multiple rounds of dialogue and identifies the business scenario as an industrial Internet of Things scenario.

[0060] The central control intelligent agent calls the parameter input intelligent agent to input the configuration according to the needs of the industrial IoT scenario, guiding the user to input parameters such as park location, equipment type, and data transmission protocol in sequence; for the "frequency band selection" drop-down parameter, it calls the remote plugin to obtain the list of currently available frequency bands; based on the "device type" entered by the user, it automatically filters out bandwidth options that do not support the type of device.

[0061] The data processing agent verifies the parameters entered by the user and finds that the "park location" format does not conform to the standard, and returns an error message. After the user corrects it, the data processing agent calls the live network resource survey interface to obtain data such as base station information and spectrum resources around the park and adds them to the parameter set.

[0062] The solution generation intelligent agent is based on the solution configuration of the industrial IoT scenario. It calls the low-latency template and the high-concurrency template, fills the processed parameters into the template and generates two activation schemes; the user selects the low-latency scheme and confirms.

[0063] The service orchestration agent parses the deterministic capability configuration corresponding to the low-latency scheme and determines the network element operation sequence: first, configure user subscription data through the UDM interface, then set QoS policies through the PCF interface, and finally configure session routing through the UPF interface; execute the above operations sequentially through the acquisition and control installation interface to complete network activation.

[0064] The content orchestration agent converts the activation results (including a list of successfully configured network elements and a test latency value of 8ms) into natural language: "Your campus 5G network is now activated, currently supporting 200 devices to connect simultaneously. The actual network latency is 8ms, meeting your needs." and provides feedback to the user.

[0065] The logical ideas behind the methods disclosed in the above embodiments can be implemented, in whole or in part, through software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs.

[0066] When computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions according to the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives (SSDs).

[0067] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0068] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0069] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0070] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments have been described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification. Although the present invention has been shown and described with reference to specific preferred embodiments, it should not be construed as limiting the invention itself. Various changes in form and detail may be made without departing from the spirit and scope of the invention as defined in the appended claims.

Claims

1. A method for enabling a customized 5G network driven by an intelligent agent, characterized in that, Includes the following steps: The central control intelligent agent engages in multiple rounds of dialogue with users to identify their network needs and the corresponding 5G customized network service scenarios. The central control intelligent agent calls the parameter input intelligent agent, which inputs the configuration according to the preset business scenario requirements, obtains the input order and format of the required parameters, and guides the user to input network activation parameters through dialogue. For dropdown parameters, call the remote plugin to obtain the data dictionary; for parameters related to previous and subsequent business processes, call the remote plugin to obtain the rules for subsequent parameters based on the entered information. The central control intelligent agent calls the data processing intelligent agent to verify the validity of the parameters obtained in step S2; If the verification passes, the data processing agent supplements the data integrity according to the network opening service input parameters, including calling the existing network resource survey capability to query the existing network data; If the verification fails, return to the central control intelligent agent to guide the user to re-enter parameters; The central control intelligent agent calls the solution generation intelligent agent. The solution generation intelligent agent is based on the solution configuration and solution template in the preset network activation solution model. It fills the parameters after verification and supplementation in step S3 into the template, and calls the large model language generation capability to generate at least one 5G customized network activation solution for user confirmation. After the user confirms the plan, the central control intelligent agent calls the service orchestration intelligent agent. The service orchestration intelligent agent calls the corresponding network activation service and network element operation commands according to the deterministic capability configuration and acquisition and control capability orchestration in the preset network activation service model to complete the 5G customized network activation. The central control agent calls the content orchestration agent, which, based on the prompts and semantic constraints configured in the knowledge base, invokes the large model's logical dialogue capabilities to organize the activation result into a user-understandable script and provides feedback to the user.

2. The method for enabling a customized 5G network driven by an intelligent agent according to claim 1, characterized in that, The solution generation agent generates at least two network activation schemes for users to choose from; the solution configuration has at least two scheme templates corresponding to each business scenario, including a high-load rate guarantee template and a mobility rate guarantee template.

3. The method for enabling a customized 5G network driven by an intelligent agent according to claim 1, characterized in that, The parameter input agent, data processing agent, content arrangement agent, solution generation agent, and service arrangement agent are designed as modular units that can be called individually or in combination.

4. The method for enabling a customized 5G network driven by an intelligent agent according to claim 1, characterized in that, The network activation scheme model extracts upper-layer application scenarios from top to bottom to form the smallest scheme unit, and the network activation service model summarizes network element capabilities from bottom to top to form a 5G customized network activation service unit.

5. A 5G customized network activation system, characterized in that, include: The central control intelligent agent is used to engage in multi-turn dialogues with users to identify network demand intentions and business scenarios, and to orchestrate and schedule various unit-based intelligent agents and external capabilities. The unitized intelligent agent set includes parameter input intelligent agent, data processing intelligent agent, content arrangement intelligent agent, solution generation intelligent agent and service arrangement intelligent agent, which respectively perform parameter input, data verification and supplementation, result script organization, activation solution generation and network activation service invocation; The network activation solution model module stores the configuration of requirement input, configuration of existing network resource survey, and solution configuration. The solution configuration includes the relationship between business scenarios and solution templates. The network activation service model module stores deterministic capability configuration and acquisition and control capability orchestration, wherein the acquisition and control capability orchestration includes network element operation commands, execution order and required parameters; The capability pool includes a knowledge base, remote plugins, acquisition and control installation interfaces, acquisition and control disassembly interfaces, DCOOS platform APIs, and resource sharing query interfaces, providing data support and interface call capabilities for each intelligent agent.

6. A 5G customized network activation system according to claim 5, characterized in that, The parameter input intelligent agent includes a requirement input configuration unit, which is used to store the parameter input order, format and dialogue skills corresponding to each business scenario. The parameter input intelligent agent guides the user to input parameters through this unit. The data processing agent includes a legality verification unit and a data supplementation unit; the legality verification unit verifies the legality of the parameters, and the data supplementation unit calls the existing network resource survey interface to obtain existing network data to supplement the completeness of the parameters.

7. A 5G customized network activation system according to claim 5, characterized in that, The solution generation agent includes a solution configuration unit and a solution template library; the solution template library stores high-load rate guarantee templates, mobility rate guarantee templates, and ULCL enhanced templates.

8. A 5G customized network activation system according to claim 5, characterized in that, The service orchestration agent includes a deterministic capability configuration unit and a data acquisition and control capability invocation unit; the deterministic capability configuration unit stores the service invocation order and required parameters corresponding to the network activation scheme, and the data acquisition and control capability invocation unit invokes the data acquisition and control interface to perform network element operations.

9. An electronic device, characterized in that, The device includes: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, it implements the agent-driven 5G customized network activation method as described in any one of claims 1 to 4.

10. A computer-readable storage medium, characterized in that, The storage medium stores at least one executable instruction, which, when executed on an electronic device, causes the electronic device to perform the agent-driven 5G customized network activation method as described in any one of claims 1 to 4.

Images