Adaptive Time Delay Estimation Based on Signal Preprocessing and Fourth-Order Cumulant

To address the problem that the traditional generalized cross-correlation (GCC) method has poor delay estimation accuracy in the low signal-to-noise ratio (SNR) environment or complex noise background, an adaptive time delay estimation algorithm based on signal preprocessing and fourth-order cumulant is proposed. We first preprocess the noisy signal using singular value decomposition and wavelet denoising. Next, we use an improved variable step-size least mean square algorithm based on multi-scale wavelet transform to iteratively operate on the one-dimensional slice of fourth-order cumulant. Finally, we derive the time delay from the peak offset of the filter weight coefficient. The simulation results show that the proposed method outperforms the GCC method and the fourth-order cumulant GCC method in Gaussian non-correlated noise, Gaussian correlated noise, and non-Gaussian noise backgrounds, achieving relatively accurate estimation results even in a low SNR environment. This technique could offer a new approach for detecting weak signals and passively locating small targets in ocean exploration.