Integration of track-to-track fusion into multiple-modely filter

Surveillance in air traffic control (ATC) uses different kinds of sensors, for instance radar, ADS-B or multi-Iateration sensors (MLAT). These sensors might be either pre-tracked individually or already combined e.g. in an SMGCS tracking system that potentially inhibits the use of Kalman filter based tracking methods in multi-sensor data fusion (MSDF). This paper addresses this particular problem in MSDF for air traffic control. We present a method that aims at combining Kalman based filtering methods with other methods such as covariance intersection on a very deep level. An implementation of this method is integrated into the PHOENIX system developed at DFS and some results will be presented.