Probability Field Localization Algorithm for Wireless Sensor Network

Sensor nodes localization is an essential requirement for many applications in wireless sensor networks. The most straightforward approach is to equip GPS receivers or other sophisticated position sensors to every node. This could be prohibitively expensive if the system consists of hundred of nodes. In this paper, we present a low cost algorithm called probability field localization algorithm, which uses a pedometer and neighbor nodes RSSI measurements to determine the nodes location. In this approach, the location of a node is derived by constructing the probability field of the sensor´s position. The simulation results show that the algorithm works well for both sparse and dense networks. Furthermore, it can achieve comparable or better localization than some existing methods. An experimental system has been setup and the results show reasonably good performance