A LEAST-SQUARES ALGORITHM FOR MULTIPATH TIME-DELAY ESTIMATION

We consider the problem of estimating the arrival times of overlapping ocean-acoustic signals from a noisy received waveform that consists o f attenuated and delayed replicas of a known transient signal. We assu me that the transmitted signal and the number of paths in the multipat h environment are known and develop an algorithm that gives least-squa res (LS) estimates of the amplitude and time delay of each path. Direc t computation of the LS estimates would involve minimization of a high ly oscillatory error function. By allowing the amplitudes to be comple x valued, a much smoother error function that is easier to minimize us ing gradient-based techniques is obtained. Using this property and the knowledge (derived from the data) of the spacing between adjacent min ima in the actual LS error function, an efficient algorithm is devised . The algorithm is a function of a data-dependent parameter, and we gi ve rules for choosing this parameter. The algorithm is demonstrated on a broad-band signal, using simulated data. The proposed method is sho wn to achieve the Cramer-Rao lower bound over a wide range of SNR's. C omparisons are made with alternating projection (AP) and estimate maxi mize (EM) algorithms.