Tracking of non-stationary biological signals with utilizing of adaptive filtering method

This paper presents an approach intended for tracking of biological non-stationary signals. The proposed approach utilizes a Kalman filter autoregressive model together with a method for estimation of covariance matrices of the uncorrelated process noise and measurement noise. The method was tested in simulations, where the ability of tracking of a class of time varying autoregressive processes was the subject of our interest. The obtained results are promising in the meaning that the suggested algorithm is suitable to track the time varying autoregressive processes with sufficient accuracy.