Performance Monitoring and Prediction for Active Management of Distributed Sensors Fusion in Target Tracking

We propose to continue our Phase I efforts by demonstrating the benefits of active management of distributed sensors and their tracking and fusion algorithms using on-line performance monitoring and prediction in Phase II. A bottom-up approach is set forth along the signal and data processing chain and sensor and network hierarchy. An extended target model will be used to facilitate ground target tracking with road constraining and feature aiding. High resolution range-Doppler processing of target returns will permit simultaneous measurement of target ID and kinematic state (i.e., position, velocity, and turn rate). On-line performance monitoring of tracking filters and fusion rules will enable adaptive tuning and active management of algorithms so as to maximize the overall performance for the given operating conditions. This will involve probabilistic vs. evidential combining of dissimilar features to assist target ID and data association as well as track fusion with memory vs. memoryless of distributed sensors. Furthermore, a local cooperative resource management policy will complement a network-wise competitive resource management strategy. The former is for specific formation coordination while the latter attempts to obtain balanced sensor-target assignments using the game theory to achieve mission success and long-term survivability. In Phase II, the computational algorithms developed in Phase I will be implemented into well-structured software tools as part of the Phase II deliverable and a “marketable” product to pursue after Phase II. At the same time, we will arrange possible tests of the proposed technology with operational tracker/fuser of a third party for demo as initial steps toward technology transition.