Underwater acoustic positioning in dense multipath channels using a 2-D wideband sparse array

In this study, we propose a new method to locate a mobile transmitter in dense multipath environment. In principle it is based on the detection of the first arrival from the beamformer output but the processing time is much reduced by estimating the positioning parameters in two stages. In the first stage, it acquires a coarse timing of the first arrival from the incoherently combined array signal and then gives the candidate angles of arrival (AOA) by the narrowband beamforming. In the second stage, it gives the precise time delay and AOA by evaluating the cross correlation between the coherently combined array signal and the probe signal over the parameter values obtained from the previous stage. Because it requires that the cross correlation be computed over the reduced parameter space obtained from the first stage, the computational complexity is much less than the wideband beamforming based method especially for 2-D array processing which has an additional AOA parameter. The performance of the proposed method is evaluated using the experimental data gathered in a very reverberant water tank using a 4 element 2-D receiver array whose sensor spacing is larger than the carrier wave length. The experimental results show that the proposed method locates the transmitter with less positioning errors than the conventional time of arrival (TOA) based positioning method.