A distributed algorithm for range-based localization in sparse wireless networks

Wireless nodes localization has become a major research theme in recent years. Distributed and collaborative solutions are particularly interesting since the nodes can compute their own localisation without depending on external computational resources. In this paper, we present the context of distributed and cooperative localisation and the principles of a distributed Force-Based localisation algorithm. We then present an original algorithm: UWL (Uncertainty Weighted Localization). A performance evaluation of UWL has been conducted by simulation and the results are compared to the well-known PPE algorithm. The results show that UWL performs better than PPE on a scenario where a comparison with PPE is possible. On two others scenarios where the number of anchors is limited, the localisation accuracy depends on the network density; With a medium connectivity, the localisation accuracy is below 1m for 99% of the nodes, even though none of them is within range of more than 1 anchor.