Distributed Scheduling for Cooperative Navigation Based on Uncertainty Evolution

Abstract

Cooperative navigation is a promising solution for many emerging applications of the Internet of Robotic Things to provide accurate location information via relative measurement, information exchange, and information fusion. However, these operations cause large communication overhead and thus are impractical for applications in resource-constrained environment. To alleviate this problem, a distributed scheduling algorithm based on uncertainty evolution under the framework of belief propagation (BP) is proposed in this paper. Specifically, we first derive a computationally efficient upper bound to characterize the approximated reduction of position uncertainty through cooperation. Then, a novel distributed cooperative navigation algorithm with candidate selection scheme is proposed for robots to guarantee satisfactory performance while reducing communication rate. Numerical simulations show that the proposed algorithm can achieve higher navigation accuracy and more efficient communication with limited on-board resources.