Author_Institution :
DSP Group, Nat. Instrum. Corp., Austin, TX, USA
Abstract :
Using the orthogonal-like Gabor expansion, the authors decompose the Wigner-Ville distribution (WVD) to a linear combination of localized and symmetric functions WVDh, h´ (t, w), the WVD of the Gabor elementary functions, hm, n(t) and hm´n´(t). Since the influence of the WVDh, h´(t, w) to the useful properties is inversely proportional to the distance between hm, n (t) and hm´, n´t, the WVDh, h´(t, w) are further grouped as a series of the function Pd(t, w). The authors name the resulting representation the time-frequency distribution series (TPDS) (also known as the Gabor spectrogram in industry). The TFDSD consists of up to a Dth order Pd(t, w). While TFDS0(t, w)=P0(t, w) is similar to the spectrogram, TFDS∞(t, w) converges to the WVDs (t, w). Numerical simulations demonstrate that adjusting the order D of the TFDS, one could effectively balance the cross-term interference and the useful properties
Keywords :
series (mathematics); signal processing; statistical analysis; time-frequency analysis; Gabor elementary functions; Gabor spectrogram; Wigner-Ville distribution; cross-term interference; linear combination; localized functions; orthogonal-like Gabor expansion; representation; series; symmetric functions; time-frequency distribution series; Digital signal processing; Fourier transforms; Frequency shift keying; Interference; Numerical simulation; Phase shift keying; Shape; Signal processing; Spectrogram; Time frequency analysis;