شناسایی مدولاسیون تطبیقی بر اساس الگوریتم های تکاملی

Nowadays, the development of classification algorithms gives the ability to improve Automatic Modulation Recognition (AMR) effectively. This paper presents a novel modulation recognition algorithm based on clustering approach. Generally, we aim to distinguish multicarrier modulation OFDM from single-carrier modulations. In this regard, two statistics of the amplitude of the received signal are calculated at the output of a quadrature mixer as key features. The extracted features of training data points are submitted to the clustering algorithm, then, centroids for single-carrier and multicarrier modulations are assessed. Afterwards, each point of testing dataset is dedicated to its nearest centroid based on Euclidian distance and the recognition is accomplished. Simulation results demonstrate that the algorithm is beneficial in a wide range from low to high SNRs.