基于卷积盲分离的强干扰下通信信号分离算法

针对目前通信辐射源个体识别算法在实际试验中由于各类干扰信号和多径衰落导致识别率较低的问题,提出一种用于识别算法前端的信号分离算法,可有效地减少其他电磁信号对于识别算法输入信号的影响,从而提高在复杂电磁环境中通信辐射源个体识别的正确识别率。该算法将灾变策略和搜索状态的自适应引入量子粒子群算法,通过对混合信号的联合对角化从截获的观测信号中提取出目标通信辐射源的有用信号。为了更加系统、直观地衡量算法的分离效果,提出分离熵来量化算法的整体性能。仿真结果表明,该分离算法可以把目标通信辐射源的有用信号从复杂电磁环境中提取出来,从而提高通信辐射源个体识别在复杂电磁环境中的正确识别率,具有较好的可行性和有效性。

Radio signal separation algorithm under strong interference based on convolutive blind separation

Aiming at the low accurate identification rate of the radio transmitter source individual identification problem caused by various jamming signals and multi-path fading in actual experiments, this paper proposes an improved blind source separation algorithm, which is used before classification algorithm and can decrease influence to input signals caused by other electromagnetic signals, thus the algorithm can be applied to increasing the accuracy rate of radio transmitter individual identification in complicated electromagnetic environment. The algorithm uses the improved quantum particle swarm optimization which brings in the cataclysm strategy and self-adaption of the seek condition for the joint diagonalization, through which can separate the objective radio source signal from the intercepted observation signals. In order to evaluate the separating effect more systematically and intuitively, the separating entropy used to quantify the integral performance of the algorithm is defined. The simulation result demonstrates that the algorithm can separate the objective communication signal from the mixed signals in complicated electromagnetic environment, which enhances the anti-jamming performance of the radio transmitter source individual identification. Therefore, the proposed algorithm is of practicability and effectiveness.