On the convergence of the estimated posterior density weighted algorithm

Sufficient conditions are established for the convergence of the algorithm when applied to the problem of radar imaging as suggested by D.L. Snyder and H.J. Whitehouse (1985), and the problem of enhancement of nonstationary spectrum estimators based on L. Cohen's (1966) class of time-frequency distributions, as proposed by L.B. White (1989). In the latter case, a regularization step must be included to ensure good behavior. The sufficient conditions may always be met in practice.<
>