Prediction of Complex-Valued Signals by Using Complex-Valued LMK Algorithm

Abstract

In this study, the adaptive prediction of complex valued signals has been realized by using the complex-valued least mean Kurtosis (CLMK) algorithm. The prediction performance of the CLMK algorithm has been evaluated on three benchmark complex-valued signals and a complex-valued real-world radar data by comparing with the traditional least mean square (CLMS) algorithm. Simulation results have verified that the CLMK algorithm outperforms the CLMS algorithm.