Comparison of Clustering Algorithms for Recognition of Radio Communication Signals based on the HOS

Milad Azarbad, Ataollah Ebrahimzade, and Mehdi Yousefi


Pattern recognition and classification, Fuzzy clustering techniques, Higher-order moments and cumulants


Automatic recognition of digital radio communication signals plays an important role in various applications. This paper introduces a comparative study of clustering algorithms on clustering of the digital modulated communication signals. We propose an efficient pattern recognition system for identification of digital communication signals. In this technique a suitable combination of the higher order moments (up to eighth) and higher order cummulants (up to eighth) and spectral characteristics are proposed as the effective features. Two different clustering algorithms are used for classification of the digital communication signals. The most important clustering techniques are Fuzzy C-means (FCM) and Subtractive clustering. Simulation results of this study show that clustering algorithm has very high recognition accuracy even at low levels of SNR with a little number of the features using proposed feature extraction methods.

Important Links:

Go Back