Cluster Based Iterative GPS-Free Localization for Wireless Sensor Networks

Precise location information is important in sensor networks to support a great many functions like location aided routing, tracking and other specific applications. This paper proposes a Cluster based Iterative GPS-Free Localization(CIGL) algorithm for wireless sensor network, which only depends upon the distances between the sensor nodes and their neighbors obtained by measurements like TOA. The proposed algorithm firstly selects a subset of nodes to establish Local Coordinate Systems (LCSs) on the basis of clustering results. By rotation and reflection transformation, all the LCSs converge to form the global coordinate system and complete the nodes localization in succession. The located nodes are nominated as new beacons to re-locate the remaining unknown nodes, namely expand localization coverage by iteration. Comparing with the previous approach, simulation results indicate that the proposed algorithm achieves considerable performance in terms of localization accuracy and coverage, especially when the communication range is small and the network deployment is sparse.