Performance analysis for frequency domain detection method in spatial color noise background

Underwater noise environment of small aperture array is spatial color noise background. A super-directive beamforming in frequency domain (FSD) method was proposed to solve the problem that the array gain of super-directive beamforming cannot reach maximum when there exists error in the estimated noise matrix under color noise background. The wideband received data were divided into many subbands, and noise covariance matrix in each subband was estimated, which reduces the error of estimated noise matrix. Estimated noise covariance matrix was used to de-correlate the received data. At last, spatial spectrum detector was employed to detect weak signal. Experimental results show that the detection performance of FSD is better than super-directive beamforming in time domain (TSD) and minimum variance distortionless response in frequency domain (FMVDR) of 2 dB and 2 dB respectively.