Kalman filtering (KF) is a form of recursive estimator, which has been widely applied to different applications. This paper discusses a new approach. based on a discrete event process, for tracking of threats (emitters). Pulses of a particular threat from received mixed (raw) pulse sequences of multiple threats corrupted by timing jitter and measurement noises are aimed to be tracked. Required parameters to initiate KF are taken from identification algorithm which begins before KF algorithm. Hereby, KF runs on mixed pulse train. The algorithm based on KF is adaptive to both the noise and missing pulses together with large time gaps between pulses of the threat. The algorithm is robust and efficient due to the fact that it includes various distinctive parameters of the threat such as frequency (TF), pulse width (DG), angle of arrival (GA) and scan patterns (TO).