A TVAR particle filter with adaptive resampling for frequency estimation

Extraction of frequency content embedded in a signal is an very important task for signal processing applications. In this paper, we present an approach for frequency tracking of a noisy Time-series using adaptive resampling particle filtering method along with the time-varying autoregressive (TVAR) model. The adaptive resampling scheme is used to address the problem of impoverishment that usually occurred in the conventional resampling stage. Simulation results demonstrate the benefits in using the adaptive method in terms of exhibiting greater tracking results of the frequency content of the signals.