Time delay estimation using Scatter Search algorithm for underwater acoustic channel

To overcome the shortcoming of low resolution of the matched filtering method when estimating time delays, the WRELAX algorithm of the Nonlinear Least-Squares (NLS) parameters estimation problem is commonly used. However, the computing time of the WRELAX algorithm is long. In this paper, the Scatter Search (SS)-based WRELAX algorithm is applied in underwater acoustical channel computation. Multipath time delay estimation of a CW signal is calculated in Gaussian white noise channel with three SNR conditions of 20dB, 10dB, and 0 dB. Simulation results show that the SS-based WRELAX algorithm reaches a higher resolution than the conventional matched filtering method. Moreover, the SS algorithm can reduce the iteration number a lot comparing with the Gauss-Newton algorithm in the process of searching the minimum objective function value.