Cooperative node localization for tactical wireless sensor networks

This paper investigates a cooperative node localization scheme for a tactical WSN (Wireless Sensor Network) that delivers accurate location estimates using a minimum number of anchor nodes. An efficient non-linear data dimension reduction technique, the Curvilinear Component Analysis (CCA), is applied, adopting a distributed map generation algorithm to create the node locations. The proposed new approach using CCA is compared with another highly effective cooperative map scheme which employs the MDS (Multi-Dimensional Scaling) non-linear reduction method, demonstrating improved position estimate accuracy and efficiency. The strength and weakness of the two schemes are analyzed through extensive simulation experiments using tactical WSN deployment scenarios.