Generalized phase synchrony using fuzzified Johansen test

Abstract

Phase synchrony is a powerful measure of analyzing nonstationary signals. We propose a new measure of phase synchrony which is the modification of GePS based on the fuzzy hypothesis testing. Generalized phase synchrony (GePS) was proposed as a multivariate index based on the linear relationships between their instantaneous frequency laws. Assessing the phase synchrony within multivariate signals is accomplished by applying the Johansen test to their intrinsic mode functions. In this paper, we developed the fuzzy version of the Johansen test. Results on the synthetic signals demonstrate the superior performance of the proposed measure compared to the original version of GePS, especially in the case of fewer numbers of channels.