Modulation classification of PSK signals in multipath and frequency offsets

In this paper, we propose and evaluate a feature-based hierarchical modulation classification algorithm to perform classification in presence of multipath channels and frequency offsets. We use a combination of features based on second-order cyclic autocorrelation function and moments-based statistics. The proposed approach does not require a priori information regarding the multipath channel, or the pulse shape used at the transmitter. The proposed algorithm achieves very good results for the COST-207 channel models in the presence of frequency offsets even at an SNR of 2 dB.