Signal Deconvolution Using Frequency and Time Domain Magnitude Constraints

In many signal restoration problems we have very limited knowledge about the statistics required for implementation of the most common restoration techniques, for example, Wiener filtering. We do, however, possess some practical a priori knowledge about the nature of the signal we week to estimate. Because of this a priori knowledge, it may be suboptimal to use methods which make few assumptions about the nature of the ideal signal, for example, maximum entropy restoration [1]. Iterative restoration methods can be easily modified to incorporate such a priori knowledge while requiring little statistical information [2].