干扰攻击下基于MAPPO的电视频谱资源分配研究

Research on TV Spectrum Resource Allocation Based on MAPPO under Jamming Attack

将认知无线电(Cognitive Radio,CR)与能量采集(Energy Harvesting,EH)进行组合是一种提高电视系统频谱性能的有效方案,但由于CR的开放性,电视网络容易遭受干扰攻击。基于此,研究了干扰攻击下EH-CR网络的联合信道和功率分配问题,考虑网络中存在随机和反应扫描两种干扰攻击,通过信道接入和功率分配,使次用户(Secondary Users,SUs)平均吞吐量最大化。将EH-CR网络建模为部分可观测的马尔可夫决策过程,提出了一个基于多智能体近端策略优化(Multi-Agent Proximal Policy Optimization,MAPPO)的联合信道和功率分配方法。仿真结果表明,提出的方法能够显著提高干扰攻击下SUs的平均吞吐量。

Combination Cognitive Radio(CR)with Energy Harvesting(EH)is an effective solution for improving the spectral performance of TV systems,but due to the open nature of CR,TV networks are vulnerable to Jamming attacks.Based on this,the joint channel and power allocation problem of EH-CR network under Jamming attacks is studied.Considering the existence of random Jamming attacks and reactive scanning Jamming attacks in the network,the average throughput of Secondary Users(SUs)can be maximized through channel access and power allocation.The EH-CR network is modeled as a partially observable Markov decision process,and a joint channel and power allocation method based on Multi-Agent Proximal Policy Optimization(MAPPO)is proposed.Simulation results show that the proposed method can significantly improve the average throughput of SUs under Jamming attacks.