结合先验知识的通信智能抗干扰技术  

Intelligent anti-jamming technology of communication combined with prior knowledge

作　　者：刘晓明 杨春[1] 刘友江[1] 曹韬[1] LIU Xiaoming;YANG Chun;LIU Youjiang;CAO Tao(Institute of Electronic Engineering,China Academy of Engineering Physics,Mianyang Sichuan 621999,China)

机构地区：[1]中国工程物理研究院电子工程研究所,四川绵阳621999

出　　处：《太赫兹科学与电子信息学报》2024年第10期1111-1116,1141,共7页Journal of Terahertz Science and Electronic Information Technology

摘　　要：针对当前智能抗干扰技术在面对快速变化的干扰表现较差的问题,提出结合先验知识网络的新型智能抗干扰技术。首先构建先验知识网络,根据历史干扰信息实现对下一时刻干扰信息的预测,使系统更好地应对快速变化的干扰;然后利用强化学习算法实现对新的干扰规律的在线学习,使算法可以适用于干扰动态变化超出离线学习模型适应范围的场景。将所提算法与无先验知识的强化学习算法进行仿真对比,结果表明,所提算法在面对快速变化的干扰时,具有更高的决策准确率和更快的收敛速度,并对环境有较好的适应性,能够有效地进行智能抗干扰。In response to the current intelligent anti-jamming technology's poor performance against rapidly changing interference,a new type of intelligent anti-jamming technology combined with priori knowledge networks is proposed.Firstly,a priori knowledge network is constructed to predict the interference information of the next moment based on historical interference information,enabling the system to better cope with rapidly changing interference;then,reinforcement learning algorithms are employed to achieve online learning of new interference patterns,allowing the algorithm to be applicable to scenarios where the dynamic changes of interference exceed the adaptation range of offline learning models.The simulation comparison between the proposed algorithm and the reinforcement learning algorithm without prior knowledge shows that the proposed algorithm has higher decision accuracy and faster convergence speed when facing rapidly changing interference,and has better adaptability to the environment,which can effectively carry out intelligent anti-jamming.

关 键 词：无线通信 抗干扰 强化学习 智能决策 先验知识 

分 类 号：TN973.3[电子电信—信号与信息处理]