A unified practical approach to modulation classification in cognitive radio using likelihood-based techniques

The automatic classification of digital modulated signals has been subject to extensive studies over the last decade, with numerous scholarly articles and research studies published. This paper provides an insightful guidance and discussion on the most practical approaches of automatic modulation classification (AMC) in cognitive radio (CR) using likelihood based (LB) statistical tests. It also suggests a novel idea of storing the known constellation sets on the receiver side using a look-up table (LUT) to detect the transmitted replica. Relevant performance measures with simulated comparisons in flat fading additive white Gaussian noise (AWGN) channels are examined. Namely, the average likelihood ratio test (ALRT), generalized LRT (GLRT) and hybrid LRT (HLRT) are particularly illustrated using linearly phase-modulated signals such as M-ary phase shift keying (MPSK) and quadrature amplitude modulation (MQAM). When the unknown signal constellation is estimated using the maximum likelihood (ML) method, results indicate that the HLRT performs well and near optimal in most situations without extra computational burden.