Warped discrete Fourier transform: A new concept in digital signal processing

Abstract

The recently introduced concept of warped discrete Fourier transform (WDFT) [2] is reviewed and its main properties are compared with that of their well-known DFT counterparts. A number of new applications, such as frequency estimation of noise-corrupted short length sinusoids, discrete multitone transmission, and oversampled nonuniform filter banks, are described. In addition, some important design issues, such as the sensitivity of computing the inverse WDFT, are included.