Generalized time–frequency distributions for multicomponent polynomial phase signals

• Published in: Signal processing 2008-04, Vol.88 (4), p.984-1001
• Main Authors: Wang, Yong, Jiang, Yi-cheng
• Format: Article
• Language: English
• Subjects: Applied sciences; Detection, estimation, filtering, equalization, prediction; Exact sciences and technology; Information, signal and communications theory; Interference terms; Matched-phase principle; PPS; Signal and communications theory; Signal, noise; Telecommunications and information theory; Time–frequency distribution; Time–frequency resolution
• ISSN: 0165-1684; 1872-7557
• DOI: 10.1016/j.sigpro.2007.10.016

The conventional time–frequency distributions for multicomponent polynomial phase signals (PPS) generally suffer from interference terms, which will obscure the true location of the auto-components in the resulting time–frequency distributions. In this paper, three schemes for designing the generalized time–frequency distributions for multicomponent PPS based on the matched-phase principle are presented, and they are illustrated as follows: (1) the scheme based on the Wigner–Ville distribution (WVD) and the L–Wigner–Ville distribution (LWVD); (2) the scheme based on the fractional matched-phase principle; (3) the scheme based on the complex lags. The interference terms induced by the nonlinearity of the signals can be suppressed; for multicomponent signals, the CLEAN technique is adopted to filter out each component by a band-pass filter, and the interference terms between different components can be suppressed (eliminated). The new generalized time–frequency distributions are superimposed to yield a high-readability representation.