基于软件无线电的电磁干扰识别系统研究

Research on Electromagnetic Interference Recognition System Based on SDR

针对通信系统中存在的电磁干扰问题,提出一种基于软件无线电的电磁干扰自动检测识别系统。该系统能够在不需要任何先验信息的情况下,利用GNU Radio软件搭配标准化硬件HackRF One平台,对干扰信号进行实时的接收识别。对干扰信号的频率和符号速率进行估计,完成时钟同步和匹配过滤;提取高阶累积量作为特征参数结合决策树分类器,实现对6种调制类型干扰信号的分类识别,以及信号的相位校正;结合信号星座图对识别结果进行判别。仿真验证表明,当信噪比为10 dB时,信号识别率可达95%以上。

Aiming at the problem of electromagnetic interference in communication system,an automatic detection and recognition system of electromagnetic interference based on SDR is proposed.The system can receive and recognize interference signals in real time by using GNU Radio software and standardized hardware HackRF One platform without any prior information.Firstly,the frequency and symbol rate of the interference signal are estimated,and the clock synchronization and matching filtering are completed.Secondly,the higher-order cumulant is extracted as the characteristic parameter,and the decision treeclassifier is combined,the classification and recognition of six modulation types of interference signals is realized,as well as the phase correction of the signal are completed.Finally,the recognition results are judged by combining the signal constellation.The simulation verifications show that when the SNR is 10 dB,the signal recognition rate can reach more than 95%.