Blind Equalization for Classification of Digital Modulations

摘要

Blind equalization for modulation classification (MC) in a frequency selective scenario is a non-trivial problem. In this paper, a new method to mitigate the inter-symbol interference (ISI) is proposed for the blind MC, where the modified constant modulus algorithm (M-CMA) precedes a modulation specific decision directed least mean square(DDLMS) stage in a cascaded equalizer scheme. The performance of the quasi hybrid likelihood ratio test (QHLRT) modulation classifier is examined under various levels of fading severity to establish the utility of the new method in practical applications.