An improved node localization algorithm for anisotropic wireless sensor networks with holes

The node localization technique as a crucial technique that affects practicality, accuracy and effectiveness of the wireless sensor networks (WSNs). Sensor nodes are often deployed nonuniformly in anisotropic WSNs with holes in many applications. The existence of holes will affect the shortest distance between nodes and result in low accuracy of node localization. In this paper, an Extended Kalman Filter Multidimensional Scaling (EKF-MDS) localization algorithm is proposed based on Multidimensional Scaling-MAP (MDS-MAP). By exploring the virtual node, it can construct the shortest path in order to estimate the distances between unknown nodes. The extended Kalman filter (EKF) algorithm is used to refine the localized coordinates which are obtained by the MDS-MAP algorithm. Extensive simulation results demonstrate that the proposed algorithm requires fewer anchors and is exceedingly accurate and efficient and is superior to existing methods in anisotropic networks with holes.