A recognition algorithm for BPSK/QPSK signals based on generalized Pareto distribution

In this paper, an algorithm for recognizing BPSK/QPSK signal is proposed. Extreme value theory is introduced, and an exceedance sequence is built from the modified spectrum obtained by the squaring operation of the observed signal, and the problem of recognizing the BPSK/QPSK signal is transformed to the problem of whether the distribution of the exceedance sequence approximately follows its standard generalized Pareto distribution or not. Finally, the Kolmogorov — Smirnov test is applied to perform the goodness-of-fit test of the generalized Pareto distribution of the exceedance sequence to complete the task of BPSK/QPSK signal recognition. Computer simulation results show that the recognition performance is good when the signal-to-noise ratio is moderate and does not require prior information such as noise variance.