Non-stationary signal processing using time-frequency filter banks with applications

We present a new approach to the design of time-frequency (TF) filter banks for non-stationary noisy signals. In this work, we concentrate on multicomponent signals represented by the minimum cross-entropy TF distribution (MCE-TFD). The proposed method is based on an array of time-varying filters, each filter processes one component of the signal according to its specific TF support. The output of the filtering algorithm is a set of the signal components. Two applications of the proposed algorithm are considered. The first application is of interference mitigation of continuous phase modulation (CPM) signals which have been extensively used in recent years, especially for wireless communications. The second application is cancellation of wideband interference in direct-sequence code division multiple access (DS-CDMA) communications, based on the MCE-TFD and the filter banks approach.