Acoustical array data from the Shallow Water Acoustics experiment was processed to show the feasibility of broadband mode decomposition as a preprocessing method to reduce the effective channel delay spread and concentrate received signal energy in a small number of independent channels. The data were collected by a vertical array designed at the Woods Hole Oceanographic Institution. Phase-shift Keying (PSK) m-sequence modulated signals with different carrier frequencies were transmitted at a distance 19.2km from the array. Even during a strong internal waves activity a low bit error rate was achieved.

Abstract-Acoustical array data from the 2006 Shallow Water Experiment (SW06) was analyzed to show the feasibility of broadband mode decomposition as a preprocessing method to reduce the effective channel delay spread and concentrate received signal energy in a small number of independent channels. The data were collected by a vertical array, which spans the water column from 12-m depth to the bottom in shallow water 80 m in depth. Binary -sequence data were used to phase-shift-keyed (PSK) modulate signals with different carrier frequencies. No error correction coding was used. The received signals were processed by a system that does not use training or pilot signals. Signals received both during periods of ordinary internal wave activity and during a period with unusually strong internal wave solitons were processed and analyzed. Different broadband mode-filtering methods were analyzed and tested. Broadband mode filtering decomposed the received signal into a number of independent signals with a reduced delay spread. The analysis of signals from the output of mode filters shows that even a simple demodulator can achieve a low bit error rate (BER) at a distance 19.2 km.Index Terms-Signal processing, underwater acoustic arrays, underwater acoustic communications.

Mode filtering is most commonly implemented using the sampled mode shapes or pseudoinverse algorithms. Buck et al. ͓J. Acoust. Soc. Am. 103, 1813-1824 ͑1998͔͒ placed these techniques in the context of a broader maximum a posteriori ͑MAP͒ framework. However, the MAP algorithm requires that the signal and noise statistics be known a priori. Adaptive array processing algorithms are candidates for improving performance without the need for a priori signal and noise statistics. A variant of the physically constrained, maximum likelihood ͑PCML͒ algorithm ͓A. L. Kraay and A. B. Baggeroer, IEEE Trans. Signal Process. 55, 4048-4063 ͑2007͔͒ is developed for mode filtering that achieves the same performance as the MAP mode filter yet does not need a priori knowledge of the signal and noise statistics. The central innovation of this adaptive mode filter is that the received signal's sample covariance matrix, as estimated by the algorithm, is constrained to be that which can be physically realized given a modal propagation model and an appropriate noise model. Shallow water simulation results are presented showing the benefit of using the PCML method in adaptive mode filtering.