Multi-agent Architecture for Robust Adaptive Decision-aiding (MARAD)

Today's ground vehicles primarily utilize the NAVSTAR Global Positioning System (GPS) for positioning and navigation (POS/NAV) which produces moderately accurate results. High accuracy POS/NAV is achieved is achieved using integrated systems that arelarge, heavy, and prohibitively expensive for large-scale deployment. Modern battlefield scenarios would greatly benefit if all ground vehicles could determine their position with high accuracy. There is also a growing non-military demand for vehiclenavigation systems that are highly accurate and can operate when GPS signals are obscured. A system comprised of a GPS receiver augmented with low-cost sensors that individually yield only moderately accurate data when integrated together produces ahighly accurate result. In our Phase I effort we developed a system that uses inexpensive accelerometers, gyroscopes, magnetometers, and vehicle wheel sensors coupled with high fidelity models of the earth's gravity and magnetic fields. Using low-cost,high performance digital signal processors (DSPs), sophisticated mathematical filters can produce high fidelity position and heading information. T use of these passive sensors is key to providing POS/NAV solutions in cases where GPS signals areunavailable due to satellite signal availability problems, enemy jamming conditions, and obstacles such as buildings common in