RePC: A Localization Method Based on Regional Partition and Cooperation in Communication Networks

Localization is emerging as a vital functionality of various wireless applications. Existing localization methods generally partition irregular regions with huge obstacles into regular ones, and rely on at least three anchor nodes to localize unkonwn nodes in each sub-region. If there are no sufficient number of anchor nodes and exist heterogeneous nodes in a particular sub-region, the corresponding localization accuracy can be greatly decreased. In this paper, we propose a new localization method RePC, which consists of low-complexity regional partition and sub-region cooperation for accurate localization. Although regional partition is a general approach, our proposed partition method differs from existing ones by achieving a low complexity of O(n). If a sub-region includes less than three anchor nodes, RePC enables a cooperation between this sub-region and other neighboring sub-regions, which offer additional anchor nodes to improve localization accuracy, even when the anchor nodes are sparsely deployed across the whole network. Due to these features, RePC is not only energy and cost efficient, but also robust and scalable for large-scale wild environment with huge obstacles. Notably, our method only needs to know the connectivity among sensor nodes, rather than the hop distance. Extensive simulations demonstrate the superior performance of the proposed method as compared with the classic range-free localization method.