Adaptive Confidence Verification for Analysis of BPSK Signals

The existing extreme statistic-based confidence test approaches for the analysis results of binary phase-shift keying (BPSK) signals, i.e., modulation recognition and parameter estimation, do not perform well under transmission impairments, partly due to their insufficient use of extreme samples. This article proposes a copula-based algorithm for the confidence test of BPSK signal analysis by combining the block maxima-based and peak-over-the-threshold-based extreme statistics. The test statistic is developed relying on the one-class detector using the copula probability density function under the null hypothesis, and the threshold is obtained adaptively through the bootstrap algorithm. The proposed statistic is proved to be asymptotically Gaussian distributed. By comparing the statistic with such a threshold, the confidence test is performed. Extensive simulations validate the effectiveness of the proposed algorithm.