Measurements and Modeling of Effects of Out-of-Plane Reverberation on the Power Delay Profile for Underwater Acoustic Channels

• Published in: IEEE journal of oceanic engineering Vol. 40; no. 4; pp. 807 - 821
• Main Authors: Jenserud, Trond, Ivansson, Sven
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.10.2015; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Channels; Decay rate; Delay; Impulses; Inclusions; Marine; Multipath channels; Propagation; Regression analysis; Reverberation; Scattering; Surface waves; Topography; underwater acoustic communication; Underwater acoustics; Underwater communication; ANE, Norway
• ISSN: 0364-9059; 1558-1691
• DOI: 10.1109/JOE.2015.2475675

Long reverberation tails are often observed in shallow-water acoustic channel impulse responses (CIRs). Examples from three significantly different environments along the Norwegian coast are presented. It is shown that 2-D propagation modeling fails to reproduce the long tails. Nevertheless, inclusion of the impulse response tails is necessary in connection with model-based simulators of communication performance, to avoid too optimistic estimates of bit error rates, etc. This is demonstrated by simulation examples for one of the environments, comparing results for truncated and complete impulse responses. High wind speeds and Doppler spread signals indicate that surface reverberation must be taken into account for modeling the tail energy. In bistatic scenarios such as the ones considered here, backscattering computations in a single vertical 2-D plane are not sufficient. A 3-D ray-based model, Rev3D, is utilized to include effects of out-of-plane scattering from the sea bottom as well as the sea surface. The scattering-strength functions include azimuthal variation, with significantly enhanced scattering close to the forward direction. Rev3D modeling supports the hypothesis that out-of-plane scattering and reverberation are main mechanisms behind the observed continuous, exponentially decaying impulse responses. Modeling incorporates available environmental information concerning sound-speed profiles, bottom topography and type, surface wave spectra, etc. Good agreement of measured and modeled time series is obtained for two of the locations. For the remaining location, modeling correctly recovers a significantly lower decay rate of the reverberation tail in comparison to the other locations, but the modeled decay rate is somewhat too large.