Study on an Echo Detection Algorithm in the Reverberation Background

Abstract

Traditional matched filtering algorithm does not work well in the reverberation background, on the basis of nonstationarity, colority and non-gaussian of reverberation, presents gaussian mixture tvar model(GTM for short) and the model formula and parameter solving methods are derived, then establish the GTM echo detection algorithm. To describe reverberation features and filtering effect quantitatively, a reverberation features evaluation method is presented. According to analysis of measured reverberation, probability density and power spectral density curve are better fitted and effective echo detection can be done by this algorithm under the reverberation background and reverberation features is improved also.