Simulating the Effect of Uncertainty in Sensor Positions on the Accuracy of Target Localization in Wireless Sensor Networks

Wireless sensor networks have gained a significant amount of attention in recent years due to a wide range of potential applications in a variety of fields. In target localization applications, the transmission of binary detection decisions has been explored as a way to overcome bandwidth limitations and prolong sensor field life by minimizing the usage of sensor resources. Some previous research on target location estimation using binary decisions has assumed that there is no uncertainty in the positions of the sensors. In reality, many factors may contribute to some measure of sensor placement uncertainty, resulting in less than ideal sensor positioning. This paper explores what happens when detection decisions generated by sensors having uncertain positions are used in target localization as though the sensor positions are known exactly. Computer simulations were used to demonstrate that sensor position uncertainty degrades the accuracy of estimation of target location along with noise and coarseness of the sensor grid