Cooperative method for wireless sensor network localization

Abstract

In order to obtain an efficient wireless sensor network localization, several enhancements based on the decentralized approach are proposed. These features can be used in the cases when multiple distance measurements are used as input, where each node iteratively updates its estimated position using a maximum likelihood estimation method based on the previously estimated positions of its neighbors. Three novel features are introduced. First, a backbone is constructed, that is, a subset of nodes that are intermediaries between multiple beacon nodes, which guides the localization process of the other (non-backbone) nodes. Second, the space is perturbed more often during the earlier time steps to avoid reaching poor local minima in some cases regarding the localization optimization function. Third, for better localization of the non-backbone (or peripheral) nodes and avoidance of the rigidity problem, 2-hop neighboring distances are approximated. The introduced features are incorporated in a range-based algorithm that is fully distributed, shows good performance, and is scalable to arbitrary network size.