Multi-sensor tracking techniques

We propose to develop algorithms and software to fuse data from multiple sensor systems that are tracking ballistic missile threat(s) in order to provide a tactical picture of higher quality than that capable from any single sensor. We will build upon our previously developed Data Fusion Correlation Algorithm (DFCA). The DFCA is a multi-source track-to-track, multi-hypothesis data fusion system developed for missile defense applications. It uses computationally robust, efficient, and theoretically rigorous algorithms and accepts a wide variety of kinematic sensor input forms including raw contact data (statistically independent) as well as reports of track state (correlated in time). All are treated within a consistent Bayesian framework. Also, the DFCA contains algorithms to estimate covariance data from knowledge of sensor characteristics, tracking geometry, and data when absent. The primary focus of this effort is to develop algorithms to estimate spatial and temporal sensor biases, to mitigate the impact of these biases upon the data fusion picture, and to quantify the improvement upon the fused tactical picture.