MDIS: A Node Localization Algorithm Based on Multi-region Division and Similarity Matching

This paper proposes an improved localization algorithm called MDIS (Multi-Region Division In Shadow). The algorithm divides the traditional overlapping communication regions constructed by neighbor anchor nodes (nodes whose locations are known) into many subregions, and coordinates of centroid points in these subregions could be calculated. Furthermore, the algorithm establishes relation arrays of the unknown node to anchor nodes and centroid points to anchor nodes. Then the location of unknown nodes in subregions is determined by calculating the correlation coefficient of above relation arrays. And simulations demonstrate that MDIS algorithm increases the accuracy and stability compared to other range-free algorithms.