Classification of MPSK Signals through Eighth-Order Statistical Signal Processing

In this work, a new automatic classifier of MPSK signals, such as 2PSK, 4PSK, 8PSK and 16PSK, is proposed. This work is based on the application of higher-order statistical signal processing. In order to identify which specific MPSK signal has been received, a particular set of eighth-order cumulants of the complex envelope is computed and used for determining the features that identify the specific MPSK signal to recognize. It was proved that such features were independent on the amplitude of the received signal and the constellation rotation angle. The classifier validation was performed through numerical simulation using Matlab and an environment corrupted by Gaussian noise was simulated. It was determined that a correct classification can be achieved at signal-to-noise ratios around 5 dB or higher.