Optimizing Sensor Locations in a Multisensor Single-Object Tracking System

Tracking a mobile object presents many challenges, especially when the tracked object is autonomous or semiautonomous and may move unpredictably. The use of autonomous mobile sensor systems allows for greater opportunity to track the mobile object but does not always yield an estimate of the tracked object's location that minimizes the estimation error. This paper presents a methodology to optimize the sensor system locations, given a single object and a fixed number of sensor systems, to achieve a position estimate that minimizes the estimation error. The tracking stations may then be controlled to achieve and maintain this optimal position, under position constraints. The theory predicts that given n sensor systems and one object there is a sensor system configuration that will yield a position estimate that minimizes the estimation error. A mathematical basis for this theory is presented and simulation and experimental results for two and three sensor system cases are shown to illustrate the effectiveness of the theory in the laboratory.