METHOD FOR FORMING ENSEMBLES OF COMPLEX SIGNALS BASED ON A LOCAL STRUCTURAL INSTABILITY INDICATOR

Abstract

The article presents a method for forming ensembles of complex signals based on a local structural instability indicator. The proposed approach provides controlled determination of time-segmentation parameters through local structural analysis of nonstationary signals and adaptive adjustment of the segmentation scale to changes in their internal organization. The method enables alignment of the ensemble structure with the actual temporal organization of the signal, which leads to a reduction in mutual correlation between ensemble elements while preserving their structural and spectral characteristics. The effectiveness of the proposed approach is confirmed by experimental modeling for various scenarios of signal structural heterogeneity, demonstrating improved ensemble noise immunity under variable interference conditions.