Situational awareness in jammed environments using track association and motion recognition

It has been shown that sensor networks and data fusion are effective in providing an accurate operational picture, even in jammed environments. However, a weakness with methods based on data from several sensors is that there is not always data available from all the sensors all the time. In such cases fusion can not be performed with good results, leading to difficulties in detecting true objects in jammed environments. The paper presents a method that can handle time periods when only one sensor in the network can observe. The method is based on a combination of track association, with data from two sensors, and motion recognition, with data from a single sensor. The method is applied to civil air-traffic monitoring. The motion recognition considers various motion patterns of the aircraft. Track association is based on statistical distance. Motion recognition is based on Hidden Markov Models (HMM). Simulation results are presented.