2021 OES China Ocean Acoustics (COA) | 2021
Research on Frequency-domain Adaptive Line Enhancement Based on Pre-whitening Matched Filter
Abstract
A frequency-domain adaptive line enhancement based on matched filter (FDAMF) is a nonlinear filtering algorithm, which greatly improves the detection performance of matched filter in an additive white Gaussian noise. Due to strong reverberation interference, the detection performance of FDAMF drops sharply in the complex and variable ocean environment, so it is difficult to detect long-range target. In order to improve the detection performance in the complex marine environment, a frequency-domain adaptive line enhancement based on pre-whitening matched filter is proposed to whiten the colored reverberation. By theoretical and simulation analysis, it can obtain better detection probability under lower input signal to noise ratio. Testing sea trials show that the processing gain of this method is about higher than 6.4dB that of the conventional method, thus an adaptive pre-whitener based on autoregressive model used in FDAMF can suppress the reverberation effectively.