一种星地网络计算卸载和资源分配方法
By constructing a PAMDP model with parameterized actions and an HPPO reinforcement learning method with a hybrid action space, the computation offloading and resource allocation of the satellite-ground network are optimized. This solves the problems of computation offloading and resource allocation in the scenarios of LEO-GEO joint use and LEO inter-satellite cooperation, and achieves reduced system latency and high efficiency in resource allocation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- XIAN INSTITUE OF SPACE RADIO TECH
- Filing Date
- 2023-10-31
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies for computational offloading and resource allocation in satellite edge networks have failed to effectively address the challenges of computational offloading and resource allocation in dynamic and complex multi-layer satellite networks. In particular, the continuity and discreteness of access satellite handover and resource allocation have not been adequately considered in scenarios involving joint use of LEO-GEO and inter-satellite cooperation of LEO satellites.
We construct a PAMDP model with parameterized actions, combine it with the HPPO reinforcement learning method in a hybrid action space, optimize the computational offloading and resource allocation of the satellite-ground network, optimize the system latency through a Markov decision process model, and comprehensively consider satellite coverage model, wireless channel model and inter-satellite cooperation.
It effectively reduces system latency and improves the efficiency of computation offloading and resource allocation, especially in access satellite switching and resource allocation in LEO-GEO joint use and LEO inter-satellite collaboration scenarios, achieving more efficient resource sharing and service collaboration.
Smart Images
Figure CN117560055B_ABST