计算机科学
强化学习
保密
物理层
地球静止轨道
理论(学习稳定性)
计算机网络
无线
分布式计算
通信卫星
卫星
访问控制
无线网络
人为噪声
广播(网络)
前向保密
空分多址
最优化问题
梯度下降
方案(数学)
可观测性
梯度法
数学优化
作者
Po-Chang Huang,Sonia Pala,Keshav Singh,Chih-Peng Li
出处
期刊:
[Institute of Electrical and Electronics Engineers]
日期:2025-05-19
卷期号:: 1-6
被引量:1
标识
DOI:10.1109/infocomwkshps65812.2025.11153014
摘要
This paper investigates secure communication in geostationary orbit (GEO) satellite networks by integrating rate-splitting multiple access (RSMA) with reconfigurable intelligent surfaces (RIS). As physical layer security (PLS) gains prominence in wireless networks, RSMA and RIS are employed to improve the secrecy rates in a multi-user MISO system, particularly for users with the lowest secrecy rates. To address the non-convex optimization problem of beamforming, RIS phase shift design, and secure rate allocation, we adopt reinforcement learning techniques, specifically comparing the performance of twin delayed deep deterministic policy gradient (TD3) and deep deterministic policy gradient (DDPG) algorithms. TD3 demonstrates superior performance, owing to its enhanced stability and reduced overestimation bias. Additionally, the study highlights the advantages of RSMA advantages over space division multiple access (SDMA) in satellite communication, particularly in managing interference. The integration of RIS further enhances secrecy rates, demonstrating RSMA’s efficacy in non-terrestrial network scenarios.
科研通智能强力驱动
Strongly Powered by AbleSci AI