On Classifiers for Feature-Based Automatic Modulation Recognition over D-STBC Cooperative Networks

This paper addresses the problem of semi-blind automatic digital modulation recognition (AMR) for wireless relaying network. The proposed recognition approach is based on a simplified Distributed Space-Time Block Coding (D-STBC) scheme for a three-node cooperative network system incorporating Amplify-and-Forward (AaF) relaying protocol network. The aim of this work is to recognize the modulation type and order among different M-ary shift keying linear modulations. In this paper, the Feature-Based Automatic Modulation Recognition (FB-AMR) is adopted through the comparison of the three most used machine learning classifiers , namely, support-vector machine (SVM), K-nearest neighbor (KNN), and naive Bayes (NB) approaches. It is shown that the SVM achieves higher modulation identification on the entire range of signal-to-noise ratio level.