Autonomous Aerial-Ground Cooperative Navigation Based on Information-Seeking in GNSS-Denied Environments

The cooperative aerial-ground robotic team attracts increasing attention for diverse Internet of Things (IoT) applications. A navigation system is essential for the team to collectively localize itself and execute missions. In this article, we propose an autonomous information-seeking navigation method in global navigation satellite system (GNSS) denied environments. Specifically, we first provide a theoretical analysis of the proposed navigation scheme in terms of the Cramer-Rao low bound (CRLB). Then, we propose a motion strategy with a weighted position matrix, which allows the robotic team to actively adapt its position to enriching position information. Moreover, a belief propagation (BP)-based algorithm is developed for the proposed scheme to improve the localization accuracy by alleviating the accuracy degradation caused by accumulated errors in inertial measurements. Numerical simulations reveal the advantages of the proposed methods, providing meaningful insights and performance evaluations for system implementations.