D-GVIO: A Buffer-Driven and Efficient Decentralized GNSS-Visual-Inertial State Estimator for Multi-Agent Systems
Yarong Luo, Wentao Lu, Chi Guo, Ming Li
AI summary
Problem
Distributed state estimation for multi-agent systems faces prohibitive computational and memory costs, communication delays, and unreliable GNSS observations on resource-constrained platforms.
Approach
The framework uses a buffer-driven architecture to modularize state propagation and updates, employs a left-invariant extended Kalman filter for accurate delayed measurement handling, and applies an adaptive buffer-based method to dynamically filter GNSS outliers.
Key results
- Modular buffer-driven architecture reduces computational and memory burdens
- L-IEKF enables efficient, accurate re-propagation of delayed measurements without costly recomputation
- Adaptive buffer-based outlier detection dynamically culls unreliable GNSS data
- Validated efficiency and robustness on open-source and real-world multi-agent datasets
Why it matters
Provides a scalable, low-bandwidth solution for real-time cooperative localization in resource-constrained multi-agent swarms operating in dynamic or GNSS-challenged environments.
Abstract
Cooperative localization is essential for swarm ap- plications like collaborative exploration and search-and-rescue missions. However, maintaining real-time capability, robustness, and computational efficiency on resource-constrained platforms presents significant challenges. To address these challenges, we propose D-GVIO, a buffer-driven and fully decentral- ized GNSS-Visual-Inertial Odometry (GVIO) framework that leverages a novel buffering strategy to support efficient and robust distributed state estimation. The proposed framework is characterized by four core mechanisms. Firstly, through covariance segmentation, covariance intersection and buffering strategy, we modularize propagation and update steps in distributed state estimation, significantly reducing computa- tional and communication burdens. Secondly, the left-invariant extended Kalman filter (L-IEKF) is adopted for information fusion, which exhibits superior state estimation performance over the traditional extended Kalman filter (EKF) since its state transition matrix is independent of the system state. Thirdly, a buffer-based re-propagation strategy is employed to handle delayed measurements efficiently and accurately by leveraging the L-IEKF, eliminating the need for costly re-computation. Finally, an adaptive buffer-driven outlier detection method is proposed to dynamically cull GNSS outliers, enhancing robustness in GNSS-challenged environments.