针对水中小目标回波信号弱、受噪声及混响强干扰影响大而难以检测的问题,提出一种基于自相关与奇异值分解(Singular Value Decomposition,SVD)的水中小目标弱信号恒虚警检测(Constant False Alarm Rate,CFAR)方法.利用CW(Continuous Wave)信号与噪声自相关的差异性,通过接收信号自相关处理降低噪声干扰.同时,利用混响相对于目标信号能量强而奇异值大的特性,通过接收信号矩阵重构与奇异值分解并对较大奇异值置零抑制混响干扰.在此基础上,采用恒虚警检测算法,可有效检测目标信号,降低虚警概率.计算机仿真和试验数据表明,该方法可有效抑制强噪声及强混响干扰,提高信号干扰比,能够实现水中小目标弱信号检测.
A constant false alarm detection method for small objects in water
Aiming at the problem that the echo signal of small targets in water is weak and difficult to detect due to the strong interference of noise and reverberation,a constant false alarm rate(CFAR)method for weak signals of small objects in water based on autocorrelation and singular value decomposition(SVD)is proposed.Taking advantage of the difference between the CW signal and the noise autocorrelation,the noise interference is reduced by the receiving signal autocorrela-tion processing.At the same time,the strong reverberation interference,which has a higher energy and larger singular values compared to the target signal,is suppressed by reconstructing the received signal matrix and setting the larger singular val-ues to zero using SVD.The CFAR algorithm is then employed to detect the target signal and reduce the false alarm probabil-ity.Computer simulations and experimental data show that this method can effectively suppress strong noise and reverbera-tion interference,improve the signal-to-interference ratio,and detect weak signals from small targets in water.
small targets in the waterautocorrelationsingular value decompositionconstant false alarm rate
NETL
NSTL
万方数据
