RESOLUTION OF A 2-PI AMBIGUITY PROBLEM IN MULTIPLE FREQUENCY SPECTRAL ESTIMATION

An algorithm is proposed to resolve a fundamental 2 pi ambiguity problem occurring in multiple frequency spectral estimation. Given M frequencies f(m), and I separate frequency estimators with unambiguous bandwidths F-i, the ambiguity problem can be stated as solving for the f(m) given the estimator outputs, alpha(mi), (1 less than or equal to m less than or equal to M; 1 less than or equal to i less than or equal to I) where f(m) = alpha(mi) + KmiFi and K-mi is some integer. The proposed algorithm exhaustively resolves all possible alpha(mi) groupings into single frequency values using a noise insensitive technique that exchanges system bandwidth for noise protection. The correct multiple frequencies are then defined as the single frequencies that repeat a specified number of times. A complete analysis is included.