Statistical Characteristic Analysis Of Seabed Reverberation In High Resolution Sonar And Research On Anti-reverberation Method

Seabed reverberation is the main interference in detecting bottom or buried target. It is difficult to detect the small target resulting from the superposition in time domain and the similarity in frequency domain between the target echo and reverberation, so it is particularly important to research on the reverberation.The main endeavor and contributions of the dissertation are as follows: one is the statistical characteristic analysis of seabed reverberation, the other is the research of anti-reverberation method.In the case of high resolution sonar, the number of scatters in one bearing resolution cell becomes smaller, which doesn't satisfy central limit theorem, so the amplitude of seabed reverberation no longer fits to rayleigh distribution. K-distribution and generalized K-distribution are presented to describe statistical characteristic of reverberation in this paper. Firstly explanation in physical mechanism and K-distribution and generalized K-distribution's probability density function are given, and then a method of hypothesis imitation is used to analyze the reverberation's statistical characteristic. At last stability of K-distribution's parameter in different trial condition and CFAR detection are discussed. The result is validated using sea trial data that the statistical property of seabed reverberation signal is similar to K-distributed.The anti-reverberation approach includes two aspects, the first is to design better active sonar waveform, the second is to meliorate signal processing arithmetic. According to this, incoherence processing of frequency-hopping signal and the pre-whiten sub-optimum basis on the hypothesis of locally stationary of reverberation are advanced. These two methods can improve the signal-to-reverberation ratio and restrain of reverberation comparing with the conventional method by simulation and process of sea trial data.