On the edge based artifact mitigation in wavelet packet transform for enhancement of spectral estimation

Wavelet packet transform is a recent addition to the rich arsenal of the signal processing tool box. In this article, we investigate the application of wavelet packet transform as a novel spectral analysis tool. The estimator is realized by a tree structure obtained by cascading a pair of half-band high and low pass filters. The main attraction for wavelet packets (WP) is the tradeoffs they offer in terms of satisfying various performance metrics such as frequency resolution, side lobe suppression and variance of the estimated power spectral density (PSD). Furthermore, the state of art in the application of wavelet transform for spectrum estimation is carried forward by bringing in a few optimizations which correct undesirable edge based artifacts that occur in the standard implementations. The systems are evaluated through simulation studies the results of which show that the proposed wavelet based approach offers great flexibility and adaptability apart from its performances which are significantly better than Fourier based estimates.