Location discovery is one of the fundamental services in Internet of things built with wireless sensor networks (WSNs). In this paper, we propose AnSuper, a novel range-free localization algorithm that provides accurate location estimates in anisotropic WSNs. In AnSuper, every anchor first selects a set of friendly anchors to which its distance estimates can be accurately obtained. These anchors are then broadcasted to neighboring common nodes, with which they can obtain accurate distance estimates to the anchors by using a novel locality-based algorithm and consequently calculate their positions. We discuss how to suppress error accumulation in the localization process. AnSuper can handle multiple anisotropy factors, including uneven nodal distribution and irregularity of the deployment region. Simulation results show that, compared with state-of-the-art solutions, AnSuper improves localization accuracy by more then 30 %.
展开▼
