Localization error minimization method with Regulated Neighborhood Distance

In wireless ad hoc and sensor networks, localization is a fundamental service to provide location information for those nodes without GPS units. In this paper, we use a simple yet effective proximity measure, called Regulated Neighborhood Distance (RND), to measure the closeness between two neighboring nodes based on their neighbors´ numbers. Furthermore, we propose a RND-based localization algorithm, called MDS-RND, to improve localization accuracy by minimizing the sum of pairwise dissimilarity error in networks. We compare our algorithm with three popular algorithms in different network scenarios, including uniform deployment, uniform deployment with coverage hole, and non-uniform deployment. Simulation results indicate that ours can achieve better localization accuracy in all network scenarios.