Confidence Evaluation of BPSK Signal Analysis Based on EVT

Abstract

Confidence evaluation algorithms for the analysis of binary phase-shift keying (BPSK) signals are discussed. The phase of the product of the observed signal and constructed reference signal is extracted, and its spectrum is utilized to verify confidence. Two extreme statistics of the phase spectrum are developed for the confidence evaluation, which depend on goodness-of-fit tests based on the generalized extreme value distribution and generalized Pareto distribution respectively. Via extensive simulations, we verify the effectiveness of the proposed algorithms.