针对主动声纳在浅海混响背景下线性调频(linear frequency modulation,LFM)信号检测的高虚警问题,提出了基于分数阶傅里叶变换(fractional Fourier transform,FRFT)的空时预白化(space-time prewhitening,STPW)方法。该方法利用FRFT对LFM信号良好的频域聚焦性,将混响信号变换到分数阶傅里叶域进行白化处理,使混响中的线变频率变换为稳定频率,增强了混响的局部稳定性,从而改善了混响的白化效果。通过实测混响数据对所提方法进行了验证,实验结果表明,所提方法使匹配滤波检测器的统计性能在信混比(signal-to-reverberation ratio,SRR)-16 dB~0 dB范围内平均提高了13%,显著提高了主动声纳在混响背景下的检测性能。
A space-time prewhitening method(STPW) based on fractional Fourier transform(FRFT) is presented to solve the high false alarm detection problem of active sonar in shallow sea.This method transforms the reverberation to fractional Fourier domain to whiten using FRFT.The linear varying frequency in linear frequency modulation(LFM) reverberation is transformed to stationary frequency,the local stationarity of reverberation is enhanced,and the prewhitening performance is improved.Finally,experiments with real reverberation data verify that the proposed method improves the detection probability of matched filter detector by 13% averagely in the range of signal-to-reverberation ratio(SRR)-16 dB~0 dB,and the detection performance of active sonar is improved significantly.