SEVENTH ANNUAL
ASAP '99 WORKSHOP

Abstract Passive acoustic discrimination of submarines from surface clutter continues to be an important problem facing the U.S. Navy's anti-submarine warfare (ASW) community. Traditional approaches rely heavily on the operational instincts of the trained sonar analyst applied to acoustic data visualization tools in the form of bearing-time records (BTRs) and lofargrams. However, as sonar system throughput continues to increase, a premium will be placed on the development of automated feature extraction and classification algorithms that cue the analyst and enable efficient scene management. Reliable automation of surface/submerged discrimination is of paramount importance in the dense-multipath, clutter-dominated littoral environment.
Over the years, a number of investigators have suggested that the fluctuation statistics of the received field may possess some capacity to distinguish surface and submerged acoustic sources. Simanin [1] proposed a statistical test based on probabilistic models for amplitude fluctuations of ray paths of surface and submerged origin. Wagstaff [2] proposed a lofargram normalization technique based on the short-time harmonic mean of FFT magnitude, which is purported to emphasize stable submarine signatures over unstable surface clutter.
More recently, Premus [3] proposed an approach that combines the fluctuation-based philosophy with a normal mode representation of the acoustic field in a shallow waveguide to discriminate source class based on the energy in temporal fluctuations of waveguide mode amplitudes. The short time-scale scintillation of normal mode amplitudes induced by target interaction with surface waves was demonstrated to possess distinct surface/submerged discrimination potential for narrowband acoustic sources in a simulated littoral waveguide [3]. In this paper, data from the ONR-sponsored shallow water experiment, SWellEX-96, is used to validate the proposed classification methodology and illustrate the basic phenomenology. Limitations imposed by imperfect knowledge of sound speed profile and bottom characteristics, the impact of limited effective vertical aperture, and the consideration of the joint distribution of mode fluctuations across mode number and frequency will be discussed.

[1] Simanin, A.A., "Possibility of Classifying Water and Surface Rays by Amplitude Fluctuations of the Received Signal," Sov. Phys. Acoust, 36 (6), Nov.-Dec. 1990, pp. 622-624.
[2] Wagstaff, R.A., "Exploitation of Fluctuations to Enhance Target Detection and to Reduce Clutter and Background Noise in the Marine Environment," Proc. SPIE Aerospace / Defence Sensing and Controls '96 Symp., Targets and Backgrounds: Characterization and Representation II, Vol. 2742, Apr. 1996, pp. 276-283.
[3] Premus, V., "Modal Scintillation Index: A Physics-based Statistic for Acoustic Source Depth Discrimination," J. Acoust. Soc. Am., Vol. 105, No. 4, Feb. 1999.

Sponsored in part by SPAWAR, under Air Force Contract F19628-95-C-0002. Opinions, interpretations, conclusions, and recommendations are those of the authors and are not necessarily endorsed by the U.S. Air Force.

Presentation (pdf format)