Association and Fusion of Active and Passive SONAR Tracks

As torpedo countermeasures evolve, it becomes more and more difficult for a single sensor to create a situational awareness picture (SAP) suitable for offensive torpedo operations. For example, the radiated spectrum of a Nixie-type device is hard to differentiate from that of its platform, yet their reflection characteristics are quite different. In order to differentiate the decoy from the target, information from multiple detection data streams must be associated and fused. Factors such as processor limitations, condensed timeline, mode changes, quick changing platform dynamics, and hardware limitations drive the requirement to provide an accurate, quick, and confident target estimate to the tactics module so the weapon can respond appropriately and in a timely manner. For this effort, Daniel H. Wagner Associates — partnering with Raytheon BBN Technologies — will design track association and fusion algorithms using measures of dissimilarity and probability to associate and fuse tracks from different detection data streams. Key features include: tracking filter processes time-late information without additional complexity; association algorithm accepts any input (including discrete attributes) for which a Bayesian filtering model can be specified; algorithm run time depends only on track loading; and, the ability to maintain track through local track breaks.