Improved auxiliary particle filtering: Application to time-varying spectral analysis


C. Andrieu, M. Davy and A. Doucet

IEEE SSP 2001,Singapore, August 2001.

Abstract: This paper addresses optimal estimation for time varying autoregressive (TVAR) models. First, we propose a statistical model on the time evolution of the requencies, moduli and real poles instead of a standard model on the AR coefficients as it makes more sense from a physical viewpoint. Second, optimal estimation involves solving a complex optimal filtering problem which does not admit any closed-form solution. We propose a new particle filtering scheme which is an improvement over the so-called auxiliary particle filter. The hyperparameters tuning the evolution of the model parameters are also estimated on-line so as to robustify the model. Simulations demonstrate the efficiency of both our model and algorithm.