Statistically enhanced localization schemes for randomly deployed wireless sensor networks

We address the localization problem for semi-static, randomly deployed, wireless sensor networks. The influence of the underlying random topology is discussed, and based on it we derive a distance estimations function of the number of hops two nodes are separated by, and characterize the precision of these estimations. The results are applied to two well-known localization schemes (DVHop and DVDistance) and improvements in the positioning error of roughly 26% for the first algorithm and 20% for the second algorithm are obtained. The introduced overhead is kept at a minimum as the resulting protocols are simpler than the original ones and only density information needs to be broadcast to the nodes.