Node ranking in wireless sensor networks with linear topology

In wireless sensor networks with linear topology, knowing the physical order in which nodes are deployed is useful not only for the target application, but also to some network services, like routing or data aggregation. Considering the limited resources of sensor nodes, the design of autonomous protocols to find this order is a challenging topic. In this paper, we propose a distributed and iterative centroid-based algorithm to address this problem. At each iteration, the algorithm selects two virtual anchors and finds the order of a subset of nodes, placed between these two anchors. The proposed algorithm requires local node connectivity knowledge and the identifier of the first sensor node of the network, which is the only manually configured parameter. This solution, scalable and lightweight from the deployment and maintenance point of view, is shown to be robust to connectivity degradation, correctly ordering more than 95% of the nodes, even under very low connectivity conditions