Distributed node location algorithm using non-anchor node clustering

摘要

In this paper a node localization algorithm for estimating unknown node positions in wireless sensor networks is presented. The proposed algorithm is based on the distributed clustering of anchor-free nodes. In the absence of anchor nodes, clusters of these sensors are formed and relevant angular and distance measurements are taken. This is followed by two phases of nodes synchronization, in which gradual adjustments of orientation and coordinates are performed utilizing a technique similar to the time synchronization. As the final synchronization process converges, clusters coordinates are merged into a unified global coordinate system, thus completing the node localization task. Experiments show that, in the case of randomly distributed nodes, the proposed algorithm achieves good performance in the areas of improved network coverage, minimized positioning error rate, and reduced communication overhead. Data show an improvement of 51.3% in network coverage in comparison to other industry-leading clustering algorithms. In addition, when the network node density is greater than 0.9, positioning error rate is reduced by an average of 22.3%.