Collaborative Autonomous Distributed Systems (CADS)

The processing software to support Navy and Marine Corps littoral battlespace operations has not kept pace with advances in sensor and communications hardware. To address this need, IET and Oregon State University were awarded a Phase I STTR to investigate a novel distributed processing approach to sensor fusion and situation assessment. In this effort, IET designed a distributed hierarchical data fusion approach at the sensor package level that supports adaptive detection, classification, identification, and tracking of targets such as surface ships, submarines, and mines at an acceptable false alarm rate. This approach supports the control of potentially large numbers of sensor packages at the system level by representing each local situation picture generated by an individual sensor package as a Bayesian Network (BN). An evolving picture based on a BN representation allows the propagation of conditional probabilities throughout the network.