An Improved Centroid Localization Algorithm in Wireless Sensor Networks

摘要

Centroid localization algorithm is a typical range-free localization algorithm. It has the merits of low complexity and low communication overhead, but its localization accuracy is very low. Mass spring model is usually used as an optimization method in the range-based localization. It can utilize the measured distances between all neighboring nodes to reduce the localization error. In this paper, we first use the centroid algorithm for coarse positioning, and then modify the mass spring model so that it can optimize the centroid algorithm without measuring distance. The simulation results indicate that the improved centroid algorithm CMSM (Centroid and Mass Spring Model) can significantly improve the localization accuracy compared with the original centroid algorithm when the communication radius is not very big.