STUDY OF FACTORS AFFECTING PERFORMANCE OF TWO ITERATIVE DECONVOLUTION TECHNIQUES.

Abstract

The authors investigate the different factors affecting the optimization parameter ( beta ) of the Guillaume-Nahman automated regularization filter and the automated optimum compensator of S. M. Riad. It has been shown that the optimum value of beta ( beta //o//p//t) is directly proportional to the applied signal energy keeping all other factors fixed. Also, it has been shown that beta //o//p//t increases towards an upper limiting value as the noise level increases, irrespective of the noise model, and the rate of change depends on the system transfer function. It is observed that changing the type of the applied signal or type of transfer function affects beta //o//p//t considerably.