Fast implementation of time-frequency distributions

Cohen´s class of bilinear time-frequency energy distributions (TFDs) offer improved resolution over linear time-frequency representations (TFRs). Several problems with TFDs persist. Many TFDs are costly to evaluate, nonrepresentational, and not associated with low-cost signal synthesis algorithms. It is shown that a recently introduced decomposition into spectrograms (SPs) can be used to construct low-cost approximations to high-resolutions TFDs. Performance examples are provided. The approximations constructed are attractive because they can be evaluated using relatively few SPs, for which there exist widely available hardware and software packages