Improving Anchor Position Accuracy for 3-D Localization in Wireless Sensor Networks

Accuracy of ordinary sensor node localization in wireless sensor networks mainly depends on the signal parameter such as time-of-arrival and signal strength estimation errors and the accuracy of the anchor node locations. In this paper a low- complexity but efficient algorithm is derived to improve anchor location accuracy in the presence of both anchor-to-anchor distance and AOA estimates and GPS measurements. Also, a Lenvenberg-Marquardt (LM) optimization based algorithm is developed for accuracy improvement when anchor-to-anchor distance estimates and GPS measurements are provided. Further, we derive the Cramer-Rao lower bound (CRLB) to benchmark the anchor position accuracy. To our knowledge, improving anchor node location accuracy and deriving the CRLB in the presence of both GPS and anchor-to-anchor measurements in 3-D scenarios are not reported in the literature. Simulation results demonstrate that the proposed approaches can improve the anchor position accuracy substantially and that the accuracy of the two developed algorithms approaches the corresponding CRLB.