Automatic recognition of radio signals using a hybrid intelligent technique

Automatic signal type recognition plays an important role in various applications. In this paper we have proposed a pattern recognition method for identification of digital signal types. In this technique a suitable combination of the higher order moments (up to eighth) and the higher order cumulants (up to eight) and spectral features are proposed a the effective features. As the classifier we have proposed a multi-class support vectors machine (SVM) based classifier that is constructed via one-against-all method. We have examined the different kernels of SVMs and compare the performances of them for automatic digital signal type identification. Experimental results show that the Gaussian radial basis function (GRBF) kernel has better performance than other kernels. Then we have used a particle swarm optimizer for selection the parameters of the classifier. Simulation results show that the proposed identifier has very high accuracy for identification of digital signal types even at low levels of SNR.