A Model of K-G Mixed Distribution for the Reverberation of High Resolution Active Sonar in Shallow Water

Abstract

In the shallow water, when the high resolution active sonar was used to detect targets, the statistical distribution of reverberation data was seriously deviates from the Rayleigh distribution. It was difficult to designs and achieves the constant false alarm rate. By analyzing the formation of clutters in active sonar and its signal processing, the clutters (i.e., isolated reverberation) obeys K distribution, and the dispersive reverberation obeys Gamma distribution. Then, a model of K-G mixed distribution for the reverberation of active sonar in shallow water was proposed in this paper, which the value of mixing factor of the mixed distribution was based on the shape parameter of K distribution. The statistical distribution curves of the K-distribution, the Gamma distribution, the K-G mixed distribution and the real reverberation data were compared, which results showed that the K-G mixed distribution was better than the K distribution and the Gamma distribution, and the model could effectively represent the statistical characteristics of real reverberation data.