Multisensor fusion for compiling battlespace tactical picture

The paper presents a modeling and simulation approach for a multisensor data fusion system to obtain a common tactical picture in defence surveillance applications. We explore an architecture for building a more accurate tactical picture. In addition, it combines evidence to determine platform position, velocity and identity parameters. One of the advantages of our method is its ability to combine effectively the processing of position and velocity information with the processing of class (vehicle type), ID (specific vehicle information) and thread type (friend/hostile) information. We define two types of fusion algorithms. The first one is based on the nearest neighbor clustering techniques for data association to determine more accurately the position and velocity of targets. The second technique is identity fusion based on a fuzzy rule based classifier.