Satellite communication network construction method based on large language model reasoning

By introducing a large language model inference engine into the satellite network, intelligent management of the satellite network in polar regions was achieved, solving the link interruption problem caused by cross-joints and reducing transmission latency and cost.

CN121036829BActive Publication Date: 2026-05-29HANGZHOU RANYANG INFORMATION TECHNOLOGY CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HANGZHOU RANYANG INFORMATION TECHNOLOGY CO LTD
Filing Date
2025-09-08
Publication Date
2026-05-29

AI Technical Summary

Technical Problem

In polar regions, satellite networks experience periodic interruptions in inter-satellite links due to inter-satellite gaps caused by satellite orbital motion. Existing solutions that rely on ground relays suffer from increased latency and are costly.

Method used

A satellite network cognitive decision-making hub, which deploys a large language model inference engine on the ground or in space, generates a network configuration policy framework through topology prediction, path search, and policy verification sub-modules. This enables forward-looking prediction and path optimization of inter-satellite links, avoiding data transmission interruptions.

Benefits of technology

It has improved the intelligence level of network management, reduced end-to-end transmission latency, ensured the continuity and reliability of communication, and reduced reliance on ground relay facilities.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121036829B_ABST
    Figure CN121036829B_ABST
Patent Text Reader

Abstract

The application discloses a satellite communication network construction method based on large language model reasoning, relates to the technical field of satellite communication and intelligent network control, and realizes the radical change of network management from how to configure to what is achieved by deep semantic analysis of high-level task instructions through a large language model engine, converting fuzzy natural language requirements into accurate executable strategy frameworks, greatly improving the intelligent level of network management and the man-machine interaction efficiency; the space-time graph model is introduced to make forward-looking prediction on the network topology, and a comprehensive link value function and a multi-objective routing evaluation function are designed, so that the system can accurately predict and avoid network faults caused by cross joints, fundamentally solve the communication continuity problem in special areas such as the polar region, get rid of the excessive dependence on ground relay facilities, and effectively reduce the end-to-end transmission delay.
Need to check novelty before this filing date? Find Prior Art