基于RSPWVD的多分量FM信号盲分离

Multi-component FM Signal Blind Separation based on Reassign-Smooth-Pseudo-Wigner-Ville Distribution

针对多分量任意形式的调频(FM)信号,利用平滑重排伪wigner-ville分布(RSPWVD)作为时频分析工具,平滑可以抑制各信号之间的交叉项,但会引起信号项的平滑,降低时频聚集性,通过重排可以提高时频聚集性。把RSPWVD的结果作为图像,利用Viterbi算法作进一步处理,分离信号分量。仿真结果表明,在未知任何信号先验知识的条件下,可以准确分离各个信号分量,得到精确的瞬时频率(IF)估计。

Use Reassign-Smooth-Pseudo-Wigner-Ville Distribution （RSPWVD） as time-frequency analysis tool disposal multi-component signal with arbitrary form frequency modulation law. Smooth can retrain cross-term of signals, but the signal-terms be smoothed at the same time. Reassign can timefrequency concentration. Use the algorithm presented in this paper analysis image of RSPWVD. Simulation experiments indicate that the algorithm can separate the instantaneous frequency of each signal component on the condition that absence of prior information about signal.