Efficient Weighted Multidimensional Scaling for Wireless Sensor Network Localization

Localization of sensor nodes is a fundamental and important problem in wireless sensor networks. Although classical multidimensional scaling (MDS) is a computationally attractive positioning method, it is statistically inefficient and cannot be applied in partially-connected sensor networks. In this correspondence, a weighted MDS algorithm is devised to circumvent these limitations. It is proved that the estimation performance of the proposed algorithm can attain Cramer-Rao lower bound (CRLB) for sufficiently small noise conditions. Computer simulations are included to contrast the performance of the proposed algorithm with the classical MDS and distributed weighted MDS algorithms as well as CRLB.