Digital Modulation Format Recognition Based On Higher Order Statistics

For the last decades, researchers have introduced many innovational methods to solve the problem of the modulation recognition for different kinds of modulation signals. Having analyzed the existing study and inspired by top researchers' methods that have been used in feature extraction, this paper proposes a digital modulation recognition method based on higher order statistics. This method deduces character parameters and distinguishes modulation formats on the ground of higher order moments of signal envelop and higher order cumulants of complex baseband signal respectively.Compared with the former modulation recognition methods based on higher order statistics, a new method based on the fourth and sixth order cumulants of received signal for classifying digital modulation signals is proposed. Since Gaussian white noise has no influence on higher order cumulants of received signal but does on the second cumulants, we do not use second cumulants here. Therefore, the influence of gaussian white noise on the recognition parameters presented in this paper is wiped off in the theory aspect. Feasibility of the proposed method is verified by large amount of experiment results presented in this paper.Also, a modulation recognition method based on the higher order moments of digital signal envelop is deducted in detail. Simulation results of this method show that it can distinguish many more kinds of modulation signals in comparison of the former methods.