An Algorithm for Target Localization in Sensor Networks Based on Overlap Area Boundary of Sensor Detection

A novel localization algorithm for moving target in sensor networks is proposed, which is based on overlap area boundary of sensor detection. The algorithm inherits the advantage of binary sensing model that sensor nodes only need provide 0 or 1 information to represent whether detect a target or not, and improves the location precision compares to traditional centroid localization algorithm that can only locate a target position between intersection area of sensor nodes. According to the spatio-temporal continuity of the trajectory of a moving target, the algorithm determines a dynamic queue of sensor nodes those have detected the target in order, and shrinks the possible location of a target into a boundary of intersection area of all nodes in the queue. In this way, the algorithm achieves better location precision than traditional centroid localization algorithm. Results of simulation experiments prove the truth of better location precision of this algorithm. Moreover, the algorithm has the features of good expansibility and robustness by avoiding whole net time synchronization instead of local communication between nodes.