Automatic modulation classification algorithm using higher-order cumulants under real-world channel conditions

Abstract

An automatic modulation classification algorithm for application in communications via multipath fading channels, without a priori information on channel characteristics, based on normalized sixth-order cumulants is described. In comparison with existing approaches, proposed algorithm can achieve much better classification accuracy in distinguishing BPSK from complex-valued modulation techniques. Theoretical analysis is verified via extensive simulations, under real-world channel propagation conditions.