Distributed geometric distance estimation in ad hoc networks

Abstract

Distributed localization algorithms for nodes in ad hoc networks are essential for many applications. A major task when localizing nodes is to accurately estimate distances. So far, distance estimation is often based on counting the minimum number of nodes on the shortest routing path (hop count) and presuming a fixed width for one hop. This is prone to error as the length of one hop can vary significantly. In this paper, a distance estimation method is proposed, which relies on the number of shared communication neighbors and applies geometric properties to the network structure. It is shown that the geometric approach provides reliable estimates for the distance between any two adjacent nodes in a network. Experiments reveal that the estimation has less relative percentage error compared to a hop based algorithm in networks with different node distributions. © 2012 Springer-Verlag.