Dynamic Multidimensional Scaling for Low-Complexity Mobile Network Tracking
Cooperative localization of mobile sensor networks is a fundamental problem which becomes challenging for anchor-less networks where there is no pre-existing infrastructure to rely on. Two cooperative mobile network tracking algorithms based on novel dynamic Multi-Dimensional Scaling (MDS) ideas are proposed. The algorithms are also extended to operate in partially connected networks. Compared with recently proposed algorithms based on the Extended and Unscented Kalman Filter (EKF and UKF), the proposed algorithms have a considerably lower computational complexity. Furthermore, model-independence, scalability as well as an acceptable accuracy make the authors' proposed algorithms a good choice for practical mobile network tracking.