Per-survivor processing for underwater acoustic communications with direct-sequence spread spectrum.

This paper proposes a receiver for direct-sequence spread spectrum transmissions in underwater acoustic channels, which combines a per-survivor processing (PSP) structure with sparse channel estimation. Specifically, the PSP structure establishes the trellis on the symbol level to render a small to moderate number of states, thus reducing the computational complexity. Meanwhile, the sparse channel estimation is performed on the chip level, where the orthogonal matching pursuit algorithm is used and a two-dimensional grid of path delay and Doppler scaling factor is incorporated in the dictionary construction. The effective combination of the PSP detection and sparse channel estimation achieves a good tradeoff between performance and complexity. Simulation and experiment results show that the proposed receiver outperforms the conventional RAKE receiver considerably, and most importantly, the proposed PSP receiver with an exact wideband dictionary maintains an excellent performance even for challenging underwater acoustic channels with large Doppler disparities on different paths.

Deterministic forward scatter from surface gravity waves.

Deterministic structures in sound reflected by gravity waves, such as focused arrivals and Doppler shifts, have implications for underwater acoustics and sonar, and the performance of underwater acoustic communications systems. A stationary phase analysis of the Helmholtz-Kirchhoff scattering integral yields the trajectory of focused arrivals and their relationship to the curvature of the surface wave field. Deterministic effects along paths up to 70 water depths long are observed in shallow water measurements of surface-scattered sound at the Martha's Vineyard Coastal Observatory. The arrival time and amplitude of surface-scattered pulses are reconciled with model calculations using measurements of surface waves made with an upward-looking sonar mounted mid-way along the propagation path. The root mean square difference between the modeled and observed pulse arrival amplitude and delay, respectively, normalized by the maximum range of amplitudes and delays, is found to be 0.2 or less for the observation periods analyzed. Cross-correlation coefficients for modeled and observed pulse arrival delays varied from 0.83 to 0.16 depending on surface conditions. Cross-correlation coefficients for normalized pulse energy for the same conditions were small and varied from 0.16 to 0.06. In contrast, the modeled and observed pulse arrival delay and amplitude statistics were in good agreement.

Physically constrained maximum likelihood mode filtering.

Mode filtering is most commonly implemented using the sampled mode shapes or pseudoinverse algorithms. Buck et al. [J. Acoust. Soc. Am. 103, 1813-1824 (1998)] placed these techniques in the context of a broader maximum a posteriori (MAP) 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 (PCML) algorithm [A. L. Kraay and A. B. Baggeroer, IEEE Trans. Signal Process. 55, 4048-4063 (2007)] 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.

Reflection of underwater sound from surface waves.

A tank experiment has been conducted to measure reflection of underwater sound from surface waves. Reflection from a wave crest leads to focusing and caustics and results in rapid variation in the received waveform as the surface wave moves. Theoretical results from wavefront modeling show that interference of three surface reflected eigenrays for each wave crest produces complicated interference waveforms. There is good agreement between theory and experiment even on the shadow side of caustics where there are two surface reflected arrivals but only one eigenray.

Modal processing for acoustic communications in shallow water experiment.

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 19.2 km from the array. Even during a strong internal waves activity a low bit error rate was achieved.

Performance analysis of adaptive equalization for coherent acoustic communications in the time-varying ocean environment.

Equations are derived for analyzing the performance of channel estimate based equalizers. The performance is characterized in terms of the mean squared soft decision error (sigma2(s)) of each equalizer. This error is decomposed into two components. These are the minimum achievable error (sigma2(0)) and the excess error (sigma2(e)). The former is the soft decision error that would be realized by the equalizer if the filter coefficient calculation were based upon perfect knowledge of the channel impulse response and statistics of the interfering noise field. The latter is the additional soft decision error that is realized due to errors in the estimates of these channel parameters. These expressions accurately predict the equalizer errors observed in the processing of experimental data by a channel estimate based decision feedback equalizer (DFE) and a passive time-reversal equalizer. Further expressions are presented that allow equalizer performance to be predicted given the scattering function of the acoustic channel. The analysis using these expressions yields insights into the features of surface scattering that most significantly impact equalizer performance in shallow water environments and motivates the implementation of a DFE that is robust with respect to channel estimation errors.

Modal analysis of broadband acoustic receptions at 3515-km range in the North Pacific using short-time Fourier techniques.

In 1995-1996 the Acoustic Thermometry of Ocean Climate (ATOC) experiment provided an opportunity to study long-range broadband transmissions over a series of months using mode-resolving vertical arrays. A 75-Hz source off the California coast transmitted broadband pulses to receiving arrays in the North Pacific, located at ranges of 3515 and 5171 km. This paper develops a short-time Fourier transform (STFT) processor for estimating the signals propagating in the lowest modes of the ocean waveguide and applies it to analyze data from the ATOC experiment. The STFT provides a convenient framework for examining processing issues associated with broadband signals. In particular, this paper discusses the required frequency resolution for mode estimation, analyzes the broadband performance of two standard modal beamforming algorithms, and explores the time/frequency tradeoffs inherent in broadband mode processing. Short-time Fourier analysis of the ATOC receptions at 3515 km reveals a complicated arrival structure in modes 1-10. This structure is characterized by frequency-selective fading and a high degree of temporal variability. At this range the first ten modes have equal average powers, and the magnitude-squared coherence between the modes is effectively zero. The coherence times of the peaks in the STFT mode estimates are on the order of 5.5 min. An analysis of mean arrival times yields modal dispersion curves and indicates that there are statistically significant shifts in travel time over 5 months of ATOC transmissions.