An adaptable architecture for blind modulations classification in variable SNR environments

Automatic classification of modulation type in detected signals is an intermediate step between signal detection and demodulation, and is also an essential task for an intelligent receiver in various civil and military applications. In this paper, a new blind classification method is proposed for additive white Gaussian noise (AWGN) channels with unknown or variable signal to noise ratios. The algorithm is capable to adapt to the input SNR. In this algorithm, a passive-aggressive learning algorithm is applied to high confidence classified samples in a general classifier that is trained by different SNR signals. The selection of appropriate features helps the general system to work for a set of initial samples of each class. Simulation results show that the accuracy of the proposed algorithm approaches to a well-trained system in the target SNR, even in low SNRs.