Statistical moments based classifier for MPSK signals

Abstract

The authors report a novel moments-based classifier to classify MPSK (M-ary phase shift keying) signals by using the moments of the phase utilizing the exact phase distribution. When compared with a case in which the Tikhonov function is used to approximate the asymptotic distribution of the phase, the new classifier with 1024 samples offered a 2 dB improvement. The 2 dB improvement is offered when the probability of misclassification is 0.01. Furthermore, improvement in performance can be obtained by increasing the length of the observation interval.<>