ASW Systems with Large Numbers of Advanced Autonomous Distributed Sensors

IET proposes a self-organizing approach to solving the data fusion aspects of the challenges faced in executing the Advanced Autonomous Distributed Sensors (AADS) concept, the Navy''s next-generation littoral ASW/AsuW surveillance approach. We proposeto mitigate the false alarm rate problem, the so-called large-N problem, posed by the AADS that arises through the use of a large number of small, inexpensive sensors in place of fewer, larger, more expensive sensors, by applying distributed data fusiontechnology. In our proposed approach, each sensor node communicates with neighboring nodes in terms of statistics such as likelihood functions or a posteriori probabilities conditioned only on the local data, rather than by making local decisions. Thoselocal statistics are then fused together to produce collaborative, global situation assessment and decisions, maintaining collective optimality.We propose to build upon IET''s state-of-the-art inference and decision-making solution techniques, developed to support computation with Bayesian and decision networks, into the self-organizing distributed data fusion infrastructure for the AADSdevelopment. IET has formulated its approach in concert with its research institution partner, Professor Kuo-Chu Chang of George Mason University''s Department of Systems Engineering and Operations Research (GMU-SEOR). Professor Chang is a widelyrecognized expert in data fusion and decision network technology. He has recently developed a novel solution for real-time decision network solutions. The result of this Phase I STTR effort will be mathematical solutions, algorithms and a software architecture for coordination among a large number of heterogeneous, inexpensive, multi-layered sensors to accomplish continuous surveillance of very largelittoral volumes. The results of this effort will be of commercial interest to any organization attempting to control a dynamic and ubiquitous sensor population for target tracking and identification. Some of the military markets include forceprotection, facility monitoring and mobile target tracking. Some of the commercial markets identified to date include the MicroElectroMechanical Systems (MEMS) market, microsatellites, terrestrial and extraterrestrial robotics, and UAVs. Potentialapplications include medical sensors, field service, law enforcement, forest fires, and disaster relief.