Length-based anchor-free distributed localization in a covered sensor network

Centralized localization dissipates a lot of energy in pooling the proximity information of the individual nodes. In contrast a distributed algorithm usually uses only local information at each node. Most localization algorithms available in the literature are are anchor based or they estimate the position of the sensors at the cost of losing accuracy. In this paper, we propose an anchor-free length-based distributed localization technique under a restricted model which solves a specialized case of this problem. The model assumes that all known distances between the sensors are less than the communication range of the sensors and all unknown distances are greater. We also assume that communication range is at least twice as much as the sensing range. Every point in the field of interest is assumed to be covered by some sensor. Each sensor localizes some of its neighbours with information from its neighbours. Each node communicates information for localization to its neighbours only; thus reducing traffic over the network. We also show that, the time required to localize all sensors uniquely under this model is linear in the number of sensors.