Cooperative localisation over small world WSN using optimal allocation of heterogeneous nodes

Abstract

Localisation of sensor nodes in a wireless sensor network (WSN) is an important problem in various applications like cyber-physical systems, Internet of things and context aware pervasive systems. Cooperative localisation using multidimensional scaling (MDS) has been used successfully in many localisation applications. A primary requirement in MDS is the computation of accurate distance estimates between pairs of nodes in a WSN. However, the estimated distances are erroneous in MDS especially for node pairs that are connected using multiple hops. This leads to an overall increase in error of location estimates. A recent development in social networks called small world phenomena can be introduced in a WSN leading to small world WSN (SW-WSN). SW-WSN exhibits low average path length and high average clustering coefficient and yields accurate distance estimates between pairs of nodes. In this work, a novel cooperative localisation method that uses heterogeneous nodes (H-nodes) is proposed over SW-WSN. In addition, two optimal H-node allocation methods are also developed for the cooperative localisation method. The significance of the proposed method in reducing localisation error, energy consumption, and bandwidth requirement is illustrated by simulations and extensive experiments on a real WSN testbed.