AImproving accuracy of delay estimation in wavelet denoising

For estimating time difference of the received signals from two spatial separated sensors, this paper puts forward a special generalized correlation algorithm for the estimation of time delay based on wavelet denoising according to wavelet denoising theory. Since the denoising of the traditional hard and soft threshold is insufficient, the selection of wavelet threshold is improved, a new wavelet thresholding technique is developed. This algorithm is not limited by a priori knowledge of signal and noise needed in the traditional generalized correlation method, and enable direct correlation method with less stringent signal and noise conditions. Simulation and experimental results indicate that the algorithm is still effective at low signal to noise ratio (SNR), and could improve the delay estimation accuracy.