MODULATION-BASED TIME-VARIANT FILTER DESIGN USING WIGNER DISTRIBUTION

The modulation-based time-variant filters are very useful for nonstati onary signal processing. They can be constructed by three stages: pre- modulation, time-invariant low-pass filtering and post-modulation. Wig ner distribution is a popular tool for showing the time-frequency dist ribution of nonstationary signals. The combination of these two techni ques is presented in this paper. We introduce Wigner distribution as a tool to estimate the modulation function for modulation-based filter. The Wigner distributions of modulation-based filters are discussed an d the Wigner distributions of output signals of modulation-based filte rs are also shown.