Affiliation:
1. School of Electronic Engineering Beijing University of Posts and Telecommunications Beijing China
2. Key Laboratory of Universal Wireless Communications, Ministry of Education Beijing University of Posts and Telecommunications Beijing China
Abstract
SummaryRecently, the Low Earth Orbit (LEO) satellite and Geostationary Earth Orbit (GEO) satellites share the same frequency to enhance spectrum efficiency in response to the scarcity of spectral resources. However, overlapping coverage areas emerge when LEO‐GEO and LEO‐LEO beams cover the same ground area in the GEO‐LEO co‐existing satellite system. Thus, the GEO‐LEO inter‐system interference and the LEO intra‐system interference are generated, which leads to more severe and complex interference. To solve the co‐channel interference problem without sacrificing the performance of the LEO system, a collaborative interference avoidance technology is proposed for the GEO‐LEO co‐existing satellite system. Firstly, the LEO satellite and user regrouping strategy (SURS) is employed to effectively avoid severe co‐channel interference by changing service satellites. Then, based on regrouping satellite and user pairs, a proximal policy optimization (PPO)‐based deep reinforcement learning (DRL) method is adopted to realize further continuous beam pointing optimization (BPO) of the LEO user antenna and reduce the co‐channel interference. Finally, simulation results verify that the proposed scheme can achieve the goal of interference mitigation and continuous service.
Funder
National Natural Science Foundation of China
Fundamental Research Funds for the Central Universities