Satellite communication network construction method based on big language model reasoning

Abstract

The invention relates to a satellite communication network construction method based on large language model reasoning, and belongs to the field of operation and maintenance management and control of a satellite communication system. The method comprises the steps that large language model setting is initialized; analyzing and processing the demand input by the large language model; the large language model performs multi-step decomposition on demand input; the large language model inquires and judges whether resources are called or not, the resources of the movable spot beams are called according to the inquiry result, beam adjustment instruction parameters are generated, and the satellite management and control system completes beam pointing adjustment; the large language model calls a resource allocation agent to complete allocation; the large language model calls a resource allocation agent; and the large language model issues network construction parameters to complete satellite communication network construction. According to the method, the inference capability of the LLM model is utilized, automatic construction of the satellite communication network can be realized, and the intelligent requirement of satellite communication is met.

Description

Technical Field

[0001] The present invention relates to the field of satellite communication, and particularly to a method for constructing a satellite communication network based on large language model reasoning, which can be used for the automatic and agile network construction in a geostationary satellite communication system. Background Art

[0002] In recent years, large language model (LLM) technologies such as Deepseek and Qwen series models have been widely applied in the field of communication. For example, by designing traffic characterization methods and adjusting the output structure of the model network to achieve network traffic understanding, by optimizing prompt words, network architecture, and assistant design to achieve network operation and maintenance management, and by identifying network attacks and generating disposal plans to ensure the security of the communication network, etc. The LLM has demonstrated powerful capabilities in performing complex step-by-step reasoning and solving problems, and has been used to solve various problems in terrestrial communication networks.

[0003] Currently, communication applications based on the LLM are more inclined towards network management and control, and the benefits it brings include: Firstly, it changes the interaction mode of the satellite network system, changing from the traditional full participation of users to the LLM initiating the understanding and processing of user intentions. The LLM splits user intentions and needs and assigns tasks, thereby providing personalized services. This brings system-level AI scheduling capabilities, and all services can be called as tools by the LLM model, which has the advantages of accurate invocation and fast response compared to traditional methods. With the capabilities of the LLM, the communication system can achieve functions such as intention understanding, changing the network interaction mode, improving management and control efficiency, and automated network operation, thus realizing "LLM empowering the network". Compared with terrestrial communication, the satellite communication field relies more on the operations of maintenance personnel for network management. Therefore, it is very suitable to implement related tasks based on large language models in the satellite communication field.

[0004] Taking the network construction of satellite communication as an example, when a communication task in a certain area comes, it is necessary to quickly form a satellite communication network for this task to ensure the communication network in this area. However, in the current actual satellite network management, it is still in the form of manually inputting task parameters. The disadvantages are: firstly, it has high requirements for the professionalism and reliability of personnel and cannot ensure the optimization of parameters. Secondly, the network formation process takes a long time, and it takes several minutes of operation to complete the entire process. However, satellite communication tasks may be urgent, and the network formation operation at the minute level may cause delays in communication opportunities. Therefore, to achieve a full-process closed-loop from operation and maintenance requirements to the generation and distribution of actual parameters for network construction, and to achieve the goal of automatic and fast and agile response network construction, a method for constructing a satellite communication network based on LLM model reasoning is needed. Summary of the Invention

[0005] To solve the above problems, the present invention proposes a method for constructing a satellite communication network based on large language model reasoning. In the field of operation and maintenance management and control of satellite communication, the present invention realizes the full-process closed-loop from operation and maintenance requirements to the generation and distribution of actual parameters for network construction, achieving the purpose of automated and rapid agile response network construction, and shortening the traditional manual satellite communication network construction time from the minute level to the second level.

[0006] The technical solution adopted by the present invention is as follows:

[0007] A method for constructing a satellite communication network based on large language model reasoning, comprising the following steps:

[0008] Step 1: Initialize the settings of the large language model, configure the satellite communication network construction Agent based on the large language model, and complete the construction of the chain of thought CoT for the network construction scenario;

[0009] Step 2: The user inputs satellite communication requirements through the large language model, and the large language model parses and processes the requirement input to determine whether to call the satellite communication network construction Agent;

[0010] Step 3: After the large language model supplements the necessary parameters through multiple rounds of inquiries, it decomposes the requirement input in multiple steps, and calls the resource query agent to query the beam resource situation that meets the requirement input from the situation system, and determines whether the queried beam resources meet the requirements;

[0011] Step 4: The large language model inquires and determines whether to call resources, calls the resources of the movable spot beam according to the inquiry result of the large language model, the large language model generates beam adjustment instruction parameters, and sends them to the satellite control system, and the satellite control system injects the instructions into the satellite to complete the beam pointing adjustment;

[0012] Step 5: The large language model determines the selected beam, completes the network system selection and user station type selection, and calls the resource allocation agent to complete the allocation;

[0013] Step 6: The large language model calls the resource allocation agent, and the resource allocation agent allocates beam and bandwidth resources by calling a small model through parameters;

[0014] Step 7: The large language model sends the network construction parameters to the satellite, terminal and operation and maintenance management and control system to complete the satellite communication network construction process.

[0015] Further, the specific method of step 1 is:

[0016] Initialize the basic large language model with the ability of tool call, and set the parameters of the large language model, including the temperature parameter, nucleus sampling probability threshold and maximum generation length;

[0017] Configure a satellite communication network construction Agent based on a large language model and complete the construction of the Chain of Thought (CoT) for the network construction scenario; the satellite communication network construction Agent is a multi-agent collaboration system composed of three agents, including a resource query agent, a resource allocation agent, and a network activation agent; the resource query agent completes the query function for satellite, beam, and transponder resources, the resource allocation agent completes the allocation function for satellite, beam, and transponder resources, and the network activation agent completes the planning and network activation functions for the allocated resources; the construction of the Chain of Thought (CoT) is based on system description of the large language model through prompt engineering.

[0018] Further, the specific method in step 2 is as follows:

[0019] The user uses natural language to input satellite communication requirements through the large language model. The large language model analyzes and processes the input of satellite communication requirements and determines whether the requirement analysis result is a satellite communication network construction requirement. If it is determined to be yes, step 3 is executed; if it is determined to be no, the process ends.

[0020] Further, the specific method in step 3 is as follows:

[0021] The large language model supplements the necessary parameters through multiple rounds of inquiries. The necessary parameters include bandwidth size, longitude, and latitude. The non-necessary parameters include satellite selection, beam selection, user station type selection, and network system selection; when the necessary parameter information is missing in the user requirement input, the large language model performs a multi-round inquiry process through Chain of Thought configuration until the user supplements all the necessary parameters.

[0022] The large language model invokes the satellite network construction Agent and decomposes the input of satellite communication requirements step by step, transforming the satellite network construction requirements into a step-by-step call workflow including a resource query agent, a resource allocation agent, and a network activation agent; the specific method is: the large language model queries the beam resource situation that meets the satellite communication requirements, uses tools or function calls to query the beam resources through the resource query agent, and determines whether the queried beam resources meet the requirements. If it is determined to be yes, step 5 is executed; if it is determined to be no, step 4 is executed.

[0023] Further, the specific method in step 4 is as follows:

[0024] The large language model asks the user whether to call the resources of the movable spot beam. After the user answers through the large language model, the large language model determines whether to call the resources. If it is determined to be no, the process ends; if it is determined to be yes, the large language model analyzes the context requirement parameters, generates movable spot beam adjustment instruction parameters, and sends them to the satellite control system. The satellite control system uploads the instructions to the satellite to complete the beam pointing adjustment.

[0025] Further, the specific manner of step 5 is as follows:

[0026] Select resources according to the results obtained by the resource query agent or adjust the results of the movable point beam, including two types of resources: one is the selection of satellites, beams, and transponders, and allocate the transponder frequency resources for the selected satellites and beams; the other is to select available earth stations and terminals, and perform network system selection and user station type selection based on user preferences and historical data.

[0027] Further, the specific manner of step 6 is as follows:

[0028] The large language model calls the resource allocation agent through a function and generates corresponding resource allocation parameters; the resource allocation agent calls a small model of the satellite communication frequency resource allocation method through the parameters to perform beam and bandwidth resource allocation; the small model includes a frequency allocation model and a beam allocation model.

[0029] Further, the specific manner of step 7 is as follows:

[0030] The large language model calls the network activation agent through a function, and issues the network construction parameters to the satellite, terminal, and operation and maintenance management and control system for execution to complete the construction of the satellite communication network.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] 1. In view of the characteristic that network management in the current satellite communication field relies more on the input of operation and maintenance personnel's requirements, the present invention utilizes the reasoning ability of the LLM model to propose a satellite communication network construction method based on the reasoning of the LLM model, which can meet the requirements of satellite communication intelligence.

[0033] 2. The present invention completes the construction of the satellite communication network through the large language model. In the field of operation and maintenance management and control of satellite communication, it can realize the full-process closed-loop from operation and maintenance requirements to the generation and issuance of actual parameters for network construction, achieving the purpose of automated and fast and agile response network construction, and shortening the traditional manual satellite communication network construction time from the order of minutes to the order of seconds. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a flowchart of a satellite communication network construction method based on the reasoning of the large language model in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0035] The technical solution of the present invention will be described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0036] A satellite communication network construction method based on the reasoning of the large language model, as Figure 1 shown, the method includes the following steps:

[0037] Step S1: Initialize the settings of the large language model (LLM). Initialize the basic large language model with the ability to call tools and set the parameters of the large language model, including the temperature parameter, the nucleus sampling probability threshold, and the maximum generation length. Specifically, the temperature parameter is set to 0.2. A lower temperature parameter is used to strictly control the randomness of the output and ensure the correctness of parameter generation. The nucleus sampling probability threshold is set to 0.6. By setting a lower nucleus sampling probability threshold, it is ensured that the LLM model generates stable and reliable answers. The maximum generation length is set to 4k or kept consistent with the maximum number of tokens allowed by the LLM model.

[0038] Subsequently, based on the LLM model, construct an Agent proxy for satellite communication network construction and complete the construction of the chain of thought (CoT) for the network construction scenario. The Agent proxy for satellite communication network construction is a multi-agent collaboration system composed of three agents, including a resource query agent, a resource allocation agent, and a network activation agent. The construction of the chain of thought is based on a systematic description of the LLM model using prompt engineering. A typical example of a systematic description is as follows: "- The network construction agent is used to automate the construction of the satellite communication network. The tools that can be used include the resource query agent, the resource allocation agent, and the network activation agent; - Parameters must be supplemented through multiple rounds of interrogation; function calls are not allowed in the absence of required parameters; - If there are no idle resources, ask whether to call the movable beam and adjust the beam direction, and appropriately call the satellite beam tool."

[0039] Step S2: The user inputs the satellite communication requirements using natural language through the LLM model. The LLM model parses and processes the input of the satellite communication requirements and determines whether the requirement parsing result is a satellite communication network construction requirement. If it is determined to be yes, then execute Step 3; if it is determined to be no, then end the process;

[0040] Step S3: The LLM model supplements the necessary parameters through multiple rounds of inquiries. The necessary parameters include bandwidth size, longitude, and latitude. The non-necessary parameters include satellite selection, beam selection, user station type selection, and network system selection. When the necessary parameter information is missing in the user demand input, the LLM model will perform a multi-round inquiry process through the described chain of thought configuration until the user supplements all the necessary parameters. Subsequently, the LLM model calls the satellite network construction Agent and decomposes the satellite communication demand input in multiple steps, transforming the satellite network construction demand into a step-by-step call workflow including a resource query agent, a resource allocation agent, and a network activation agent. First, to complete the network construction demand, the LLM model needs to query the beam resource situation that meets the demand, query the beam resources through the tool usage or function call (Function call) resource query agent, and determine whether the queried beam resources meet the demand. If the judgment is yes, step S5 is executed; if the judgment is no, step S4 is executed;

[0041] Step S4: The LLM model asks the user whether to call the resources of the movable spot beam. After the user answers through the LLM model, the LLM model determines whether to call the resources. If the judgment is no, the process ends; if the judgment is yes, the LLM model parses the context demand parameters, generates the movable spot beam adjustment instruction parameters, sends them to the satellite control system, and uploads the instructions to the satellite to complete the beam pointing adjustment, and then executes step S5;

[0042] Step S5: The LLM model determines the selected beam, completes the network system selection, completes the user station type selection, and calls the resource allocation agent to complete the allocation. Based on the results obtained from the resource query agent or the results of adjusting the movable spot beam, resources are selected, including two types of resources: one is the selection of satellites, beams, and transponders, and frequency resources of the transponders are allocated to the selected satellites and beams. For example, 30M available resources are divided from 100M transponder resources, and time conflicts are ensured. The other is to select available earth stations and terminals, and perform network system selection and user station type selection based on user preferences and historical data.

[0043] Step S6: The LLM model calls the resource allocation agent through a function and generates corresponding resource allocation parameters. The resource allocation agent allocates beam and bandwidth resources by calling a small model of the satellite communication frequency resource allocation method through parameters. The small models include a frequency allocation model and a beam allocation model, both of which are existing technologies and will not be elaborated here;

[0044] S7: The LLM model calls the network activation agent through a function, sends the network construction parameters to the satellite, terminal, and operation and maintenance control system for execution, and completes the satellite communication network construction process.

[0045] The present invention realizes one - key network construction for satellite communication based on the general large - model LLM and Agent proxy. In the field of operation and maintenance management and control of satellite communication, by using the large - language model to complete the construction of the satellite communication network, a full - process closed - loop from operation and maintenance requirements to the generation and distribution of actual parameters for network construction can be achieved, reaching the goal of network construction with automation and rapid and agile response, and shortening the traditional manual satellite communication network construction time from the minute level to the second level.

[0046] The technical features of the above - mentioned embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above - mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this specification.

[0047] The above - mentioned embodiments only represent several implementation manners of the present invention. Their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the invention patent should be subject to the appended claims.

Claims

1. A method for constructing a satellite communication network based on large language model reasoning, characterized in that, It includes the following steps: Step 1: Initialize the settings of the large language model, configure the satellite communication network construction Agent based on the large language model, and complete the construction of the chain of thought CoT for the network construction scenario; Step 2: The user inputs satellite communication requirements through the large language model. The large language model parses and processes the requirement input and determines whether to call the satellite communication network construction Agent; Step 3: After the large language model supplements the necessary parameters through multiple rounds of inquiries, it decomposes the requirement input in multiple steps and calls the resource query Agent to query the beam resource situation that meets the requirement input from the situation system, and determines whether the queried beam resources meet the requirements; Step 4: The large language model inquires and determines whether to call resources. According to the inquiry result of the large language model, it calls the resources of the movable point beam. The large language model generates beam adjustment instruction parameters and sends them to the satellite control system. The satellite control system injects the instructions into the satellite to complete the beam pointing adjustment; Step 5: The large language model determines the selected beam, completes the network system selection and user station type selection, and calls the resource allocation Agent to complete the allocation; Step 6: The large language model calls the resource allocation Agent, and the resource allocation Agent calls a small model through parameters to allocate beam and bandwidth resources; Step 7: The large language model sends the network construction parameters to the satellite, terminal and operation and maintenance control system to complete the satellite communication network construction process.

2. The method for constructing a satellite communication network based on large language model reasoning according to claim 1, wherein The specific method of Step 1 is: Initialize the basic large language model with the ability to call tools, and set the parameters of the large language model, including the temperature parameter, nucleus sampling probability threshold, and maximum generation length; Configure the satellite communication network construction Agent based on the large language model and complete the construction of the chain of thought CoT for the network construction scenario; the satellite communication network construction Agent is a multi-agent collaboration system composed of three agents, including a resource query Agent, a resource allocation Agent, and a network activation Agent; The resource query Agent completes the query function of satellite, beam, and transponder resources. The resource allocation Agent completes the allocation function of satellite, beam, and transponder resources. The network activation Agent completes the planning and network activation function of the allocated resources; the construction of the chain of thought CoT is based on the system description of the large language model through prompt engineering.

3. A method for constructing a satellite communication network based on large language model reasoning according to claim 1, characterized in that, The specific method of Step 2 is: The user uses natural language to input satellite communication requirements through the large language model. The large language model parses and processes the satellite communication requirement input and determines whether the requirement parsing result is a satellite communication network construction requirement. If it is determined to be yes, then execute Step 3; if it is determined to be no, then end the process.

4. A method for constructing a satellite communication network based on large language model reasoning according to claim 1, characterized in that, The specific method of Step 3 is: The large language model supplements the necessary parameters through multiple rounds of inquiries. The necessary parameters include bandwidth size, longitude, and latitude. The non-necessary parameters include satellite selection, beam selection, user station type selection, and network system selection; when the necessary parameter information is missing in the user requirement input, the large language model conducts multiple rounds of inquiry processes through the chain of thought configuration until the user supplements all the necessary parameters; The large language model makes Agent calls for satellite network construction, and performs multi-step decomposition on the input of satellite communication requirements, transforming the satellite network construction requirements into a step-by-step call workflow including a resource query agent, a resource allocation agent, and a network activation agent. The specific method is as follows: The large language model queries the beam resource situation that meets the satellite communication requirements, queries the beam resources through the use of tools or function calls to the resource query agent, and determines whether the queried beam resources meet the requirements. If the determination is yes, step 5 is executed; if the determination is no, step 4 is executed.

5. A method for constructing a satellite communication network based on large language model reasoning according to claim 1, wherein The specific method of step 4 is as follows: The large language model asks the user whether to call the resources of the movable spot beam. After the user answers through the large language model, the large language model determines whether to call the resources. If the determination is no, the process ends. If the determination is yes, the large language model parses the context requirement parameters to generate movable spot beam adjustment instruction parameters, which are sent to the satellite control system. The satellite control system uploads the instructions to the satellite to complete the beam pointing adjustment.

6. The method for constructing a satellite communication network based on large language model reasoning according to claim 1, wherein The specific method of step 5 is as follows: Resources are selected according to the results obtained by the resource query agent or the results of adjusting the movable spot beam, including two types of resources: one is the selection of satellites, beams, and transponders, and the frequency resources of the transponders are allocated to the selected satellites and beams; the other is to select available earth stations and terminals, and network system selection and user station type selection are performed based on user preferences and historical data.

7. A method for constructing a satellite communication network based on large language model reasoning according to claim 1, characterized in that, The specific method of step 6 is as follows: The large language model calls the resource allocation agent through a function and generates corresponding resource allocation parameters; the resource allocation agent calls a small model of the satellite communication frequency resource allocation method through the parameters to perform beam and bandwidth resource allocation; the small model includes a frequency allocation model and a beam allocation model.

8. A method for constructing a satellite communication network based on large language model reasoning according to claim 1, characterized in that The specific method of step 7 is as follows: The large language model calls the network activation agent through a function, sends the network construction parameters to the satellite, terminal, and operation and maintenance control system for execution, and completes the construction of the satellite communication network.

Images