Linear Prediction with Adaptive Filters by using Clonal Selection Algorithm

In this paper, linear prediction of signals is realized with an adaptive filter structure using a clonal selection algorithm (CLONALG). The performances of gradient based least mean squares method and the clonal selection algorithm are compared. Furthermore, the effect of the adaptive filter order for proposed approach on the performance is examined and with higher order filters smaller prediction errors are obtained. Also, it is seen that the dependence of the clonal selection algorithm to the design parameters is less than the conventional methods.