Sensor selection algorithm based on weighted distance in wireless sensor networks tracking

A new sensor selection methods for bearings-only sensors, weighted-distance node selection (WNS) was proposed. Based on probability distribution function (PDF) of target estimates and sensing-model, the sensors with the minimum weighted-distance were activated in the next snapshot. Since the existing mahalanobis distance measure, though yields a good precision with low computational complexity, fits only range sensors and the algorithms based on the PDF, such as entropy method, are computationally burdensome, WNS avoids the Bayes filtering process and achieves alluring precision. Simulation results show that the computation complexity is indeed reduced while a comparable selection precision is achieved as entropy method.