Frequency-Domain Deconvolution Adaptive Noise Cancellation DOA Estimation Algorithm Based on Matched Filtering

Abstract

Aiming at the problems of low azimuth estimation accuracy and wide beamwidth of single-vector hydrophone in a low signal-to-noise ratio (SNR) environment, this article proposes an azimuth estimation algorithm based on matched filter sound pressure frequency deconvolution adaptive noise cancellation (MF-FDAC). This method uses the frequency-domain output signal of each channel of the micro-electro-mechanical system (MEMS) vector hydrophone after matched filtering and finds that the two velocity components still show the sine and cosine weighting characteristics of the sound pressure propagation direction in the expression. By convoluting the frequency-domain signal and using the residual energy difference between the output of the adaptive canceller on the target azimuth and the non-target azimuth, the target azimuth is estimated. The simulation results show that the root-mean-square error (RMSE) of the proposed algorithm is about 9° and the detection probability is 0.45 when the SNR is −10 dB. The experimental results show that the maximum error of direction-of-arrival (DOA) estimation is 2° when the SNR is −7 dB. In summary, compared with conventional beam forming (CBF), multiple signal classification (MUSIC), and other algorithms, the proposed method has higher resolution, lower azimuth estimation error, and more robust azimuth estimation ability under low SNR conditions.