A Dynamic MDS-Based Localization Algorithm for Mobile Sensor Networks

In this paper, we proposed a dynamic mobility-assisted MDS-based localization algorithms for sparse mobile sensor network. For sparse networks, the assumption of the existing MDS-based localization algorithms is not necessary valid and the network may even be nonrigid, which significantly affects the application and accuracy of the existing MDS-based algorithms. In the proposed algorithm, we utilize the mobile capability of sensors in a mobile network. By moving the sensors in a random direction and recording the distances to their neighbors during the movement, virtual nodes are added. The distances between virtual nodes and real nodes provide more information about network, which leads to significantly better localization than existing methods for sparse networks. Experiments and evaluation of the proposed algorithm are provided.