基于强化学习的信道碰撞对抗研究  

Research on Channel Collision Jamming Based on Reinforcement Learning

作　　者：谢海东 陈远清 向雪霜 XIE Haidong;CHEN Yuanqing;XIANG Xueshuang(Qian Xuesen Laboratory,Chinese Academy of Space Technology,Beijing 100190,China)

机构地区：[1]中国空间技术研究院钱学森空间技术实验室,北京100190

出　　处：《无线电通信技术》2022年第4期745-750,共6页Radio Communications Technology

基　　金：国家自然科学基金(12004422);北京市科技新星计划(Z191100001119129)。

摘　　要：随着未来物联网的发展,智能家居和智能仓储等场景下,高并发通信需求快速增加。各种多样化的协议满足了低成本、低功耗、易部署的需求。但其遵循的尽力而为的思想无法确保高并发场景中,当存在非合作或恶意干扰信号时,通信体系仍然能够减少碰撞和可靠安全的运行。从现有防碰撞策略入手,提出了基于强化学习的自适应信道干扰方法,对比已有典型方法,运用强化学习技术可以智能寻找到更低功耗与更高效能的智能干扰策略;进而考虑具备一定规模的高并发场景,运用所提出的自适应信道干扰方法实现了有效的信道碰撞阻塞,并且发现当去掉干扰后阻塞现象会在一定时间内持续保持,表明已有防碰撞策略不具备抗干扰能力。With the development of the internet of things(IoT),the demand for high concurrent communication will increase rapidly in scenarios such as smart home and smart storage.Diversified protocols can meet the needs of low cost,low power consumption and easy deployment.However,the best effort idea cannot ensure reliable and safe operation of communication system in high concurrency scenarios,especially when there are non-cooperative or malicious signals.Starting with existing anti-collision strategies,this work proposes an adaptive channel jamming method based on reinforcement learning.Compared with existing typical methods,the use of reinforcement learning technology can find an intelligent jamming strategy with lower power and higher efficiency.Furthermore,considering the high concurrency scenario with a certain scale,the adaptive channel jamming method proposed in this paper is used to realize effective channel collision blocking,and it is found that the blocking phenomenon will continue for a long time after removing the jamming,indicating that the existing anti-collision strategy does not have anti-jamming ability.

关 键 词：无线并发通信 防碰撞策略 通信干扰与抗干扰 强化学习 

分 类 号：TN919.23[电子电信—通信与信息系统]