Comparison of frequency estimators for underwater acoustic data

Abstract

A number of high-accuracy frequency estimators have been critically compared using simulated and real passive sonar data. Techniques considered are the standard Prony method, the extended Prony method of Marple, the Kumaresan–Prony algorithm, the Pisarenko algorithm, the maximum likelihood approach, and a phase interpolation method developed by the authors. The latter two techniques appear to offer the most promise for the analysis of sonar data.