Enabling Netted Sensor Fusion for Anti-Submarine Warfare in Uncertain and Variable Environments

Daniel H. Wagner Associates proposes to develop a tool set for object location and classification, and sensor data geo-registration for full distributed platform data fusion and multi-attribute ASW scene generation. The process of accurately locating and identifying tracks and objects using multiple sensors requires that the individual sensor data be properly geo-registered prior to the data fusion, classification, and location refinement processes to avoid false classifications and misidentified threats. Proposed is an innovative method to register multiple sensor data into a high probability scene using a stochastic multi-hypothesis pairwise registration algorithm. The registered data is then post processed using a multi-hypothesis tracker to produce the most likely scenarios. Our approach to advanced geo-registration algorithms to fuse data from multiple platform sensors includes individual sensor performance, multiple attribute, and maximum likelihood optimal assignment. Once a common ASW scene is created we will classify the track with an inferential reasoning engine used to estimate target ID is based on a context dependent Bayesian Network.