Track-to-track fusion module for network-centric air surveillance applications

Track-to-track fusion is required in network-centric multisensor surveillance systems utilizing hierarchical or truly decentralized data processing architectures. FusionTracker is a software module, which performs real-time track-to-track fusion on general purpose computing platforms, and which is intended for application in systems utilizing networked air surveillance radars. It supports a few common association metrics and fusion rules, and employs the global restart approach to avoid double counting of sensor data due to repeated communication and fusion. In this paper, we describe the design of FusionTracker, we provide preliminary evaluation of its performance using both simulated and real-world air surveillance data, and we define directions of further research and development.