Analytic Signal Decomposition Using M-Channel Signal Matched Filter Bank
Geetanjali Srivastava, Shiv Dutt Joshi · 2017
Current signal decomposition and TFA methods like STFT (uses pre-defined computation window), EMD (not completely intrinsic, uses interpolator), PCA and ICA (both statistical based) have inherent limitations of decomposing signals into uncorrelated narrow bands and are not completely signal adaptive. In this paper, we propose an algorithm for a given data case based on signal matched filter banks having signal adaptive feature to decompose and analyze any time varying signal into uncorrelated narrow bands evaluated at immediate arrival of data. To achieve this, we have modified synthesis side of SMFB and used analytic counter-part of the original signal.