无线电工程2025，Vol.55Issue(4)：891-897,7.DOI:10.3969/j.issn.1003-3106.2025.04.024
基于改进贪婪策略的NGSO干扰减缓方法
NGSO Radio Astronomy Interference Mitigation Based on Improved Greedy Strategy
袁书豪 1尹良1
作者信息
- 1. 北京邮电大学 信息与通信工程学院,北京 100876
- 折叠
摘要
Abstract
In order to meet the requirements for 6G global coverage and high-rate communications,several large-scale Low Earth Orbit(LEO)satellite constellations have emerged or will take shape.Among them,LEO polar orbit constellations have serious overlapping coverage problems in middle and high latitude regions.This leads to the waste of communication resources and inter-system co-channel interference.By implementing the Beam Switch Off(BSO)strategy to turn off certain beams,it is possible to reduce the overlapping coverage of beams and save satellite energy as well.Regarding the co-frequency interference,it is necessary to optimize the selection of beams to be deactivated,and increase the weight of the interfering beams to be deactivated.A Deep Reinforcement Learning(DRL)enhanced greedy strategy is proposed for this optimization problem.In the Iridium Next constellation scenario,a realistic antenna model is used.Simulation results show that the algorithm effectively deactivates 37.4%of the beams and reduces the system's Equivalent Power Flux Density(EPFD)to not more than-194 dB(W/m2)in 20 kHz reference bandwidth for at least 98%of the time,which meets the limit requirements of the International Telecommunication Union(ITU).While implementing BSO,the inter-system interference is effectively mitigated.关键词
低轨卫星/多波束/干扰减缓/深度强化学习/等效功率通量密度Key words
LEO/multibeam/interference mitigation/DRL/EPFD分类
信息技术与安全科学引用本文复制引用
袁书豪,尹良..基于改进贪婪策略的NGSO干扰减缓方法[J].无线电工程,2025,55(4):891-897,7.
