Sea Clutter Modeling Improvement and Target Detection by Tsallis Distribution

Sea clutter is the backscattered returns from a patch of sea surface illuminated by a radar pulse and it´s one of difficult domains in radar clutter modeling. A lot of efforts have been made to fit various distributions to the observed amplitude data of sea clutter. However, the fitting of those distributions to real sea clutter data is not good, and using parameters estimated from those distributions is not very effective for detecting targets within sea clutter. This may be due to the fact that sea clutter is highly non-stationary. Tsallis distribution is one of distributions that recommend in recent year for sea clutter modeling. This distribution is obtained by maximizing the Tsallis entropy. The Tsallis entropy is a generalization of the Shanon entropy. In this paper was found two weak points, by accomplished simulation analyses: a) By reason of using small step in parameters estimation, the time of modeling is long. b) By reason of using short segment of clutter data, decrease the target detection accuracy. By using the bigger step in this paper, increase the parameter estimation quickness and by using all sea clutter data and decrease performance, was improving the target detection accuracy.