High resolution time-frequency analysis by fractional domain warping

Date of Conference: 7-11 May 2001 Conference name: 2001 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings A new algorithm is proposed to obtain very high resolution time-frequency analysis of signal components with curved time-frequency supports. The proposed algorithm is based on fractional Fourier domain warping concept introduced in this work. By integrating this warping concept to the recently developed directionally smoothed Wigner distribution algorithm [1], the high performance of that algorithm on linear, chirp-like components is extended to signal components with curved time-frequency supports. The main advantage of the algorithm is its ability to suppress not only the cross-cross terms, but also the auto-cross terms in the Wigner distribution. For a signal with N samples duration, the computational complexity of the algorithm is O(N log N) flops for each computed slice of the new time-frequency distribution.