Sensor Data Fusion

The Ballistic Missile Defense System (BMDS) relies on a geographically dispersed set of sensors operating across different frequency bands. As tracking and discrimination techniques that combine information from multiple BMDS assets are superior to techniques that rely on individual sensors alone, robust data fusion algorithms are needed to maximize performance. Data fusion algorithms that provide 3-dimensional images of targets, as well as estimates of target motion parameters and tests for rigidity, offer promising new capabilities to the BMDS. Technology Service Corporation (TSC) will develop and analyze a suite of algorithms that provide these target features in the context of a multi-sensor network of radars that are not cohered with one another and have a limited Coherent Processing Interval (CPI). The successful demonstration of such a system will provide target features and discrimination with minimal radar resources. In Phase I, TSC will investigate spectral estimation and 3-dimensional tomographic imaging techniques, as well as extend proven physics-based techniques to operate under a multi-sensor network. In Phase II, TSC will refine the algorithms to work with measured data and pursue speed improvements for real time applications.