基于CNN的雷达嵌入式通信接收技术研究

Research on Radar-embedded Communication Receiving Technology Based on CNN

雷达嵌入式通信(Radar-embedded Communication,REC)是一种在雷达后向散射回波中嵌入通信信号的新型隐蔽通信手段。为了进一步提高REC系统中通信接收机的接收性能,提出了一种基于卷积神经网络(Convolutional Neural Networks,CNN)的REC合作接收机结构———CNN接收机,给出了CNN接收机的设计原理及详细结构,并对CNN接收机的接收性能进行了仿真和分析。结果表明,相较于传统REC合作接收机,CNN接收机具有更优的接收性能,REC系统可以使用功率更低的通信信号来保证相同的通信可靠性,以此来增加抗截获性能。

Radar-embedded Communication(REC)is a new type of covert communication method in which communication signals are embedded in radar backscatter echoes.In order to further improve the receiving performance of communication receiver in REC system,a REC cooperative receiver structure based on Convolutional Neural Networks(CNN)———CNN receiver is proposed.The design principle and detailed structure of CNN receiver are given,and the receiving performance of CNN receiver is simulated and analyzed.The results show that as compared with traditional REC cooperative receiver,the CNN receiver is of better reception performance.The REC system ensures the same communication reliability by using lower power communication signals,so as to increase the antiinterception performance.