A Robust Kalman Filter for Estimatioh and Tracking of a Class of Periodic Discrete Event Processes

This paper discusses a Kalman filter approach to parameter estimation and tracking for a class of discrete event processes. The proposed estimation techniques operate on the recorded event arrival time sequence of a pulse train signal with pulse occurrence times corrupted by timing noise. In adopting a state space approach to signal modelling, a number of real-world conditions are considered and this leads to the formulation of a ICalman filter estimator that is robust to missing and false data, and to signal model mismatch.