An SNR estimation based adaptive hierarchical modulation classification method to recognize M-ary QAM and M-ary PSK signals

Abstract

Robust automatic modulation recognition techniques are in demand for software defined radios and cognitive radio receivers in general and Communication Intelligence (COMINT) applications such as signal interception for mobile communication, defense, civil authority and surveillance in particular. It is also the key for threat analysis. Many algorithms have been proposed and studied in the literature to distinguish digitally modulated signals. The approach presented in this paper gives a robust method for recognizing a subclass of digital modulation methods using SNR estimation and higher order cumulants. We have proposed a new feature to extend the hierarchical modulation classification proposed in the literature to include QPSK and 8PSK separation. Using higher order statistic based SNR estimation method at the receiver end we have given an SNR based adaptive thresholding technique which is capable of differentiating 16QAM, 32QAM, 64QAM, BPSK, QPSK, 8PSK signals with better false alarm rates compared to fixed threshold based recognition. Simulation results demonstrate that our approach is robust for all SNRs which can be estimated well at the receiver end.