Exploiting redundancy for UWB anomaly detection in infrastructure-free multi-robot relative localization.

Area of Science:

• Robotics
• Localization and Navigation

Background:

• Ultra-wideband (UWB) localization offers cost-effective and accurate positioning in Global Navigation Satellite System (GNSS)-denied environments.
• Existing research on UWB resilience primarily focuses on non-line-of-sight and multipath detection, with limited attention to multi-node relative localization disturbances.

Purpose of the Study:

• To develop an approach for detecting range anomalies in UWB ranging measurements within multi-robot cooperative localization systems.
• To enhance the resilience of relative localization in multi-robot systems against ranging disturbances.

Main Methods:

• Exploiting redundancy in relative localization by calculating each node's position using different data subsets.
• Identifying and eliminating the influence of nodes exhibiting ranging anomalies within the cooperative localization framework.

Main Results:

• The proposed method effectively identifies nodes with ranging anomalies, such as those caused by timing errors or hardware malfunctions.
• The approach leads to a more resilient cooperative localization framework.

Conclusions:

• The developed method enhances the robustness of UWB-based cooperative localization in multi-robot systems.
• The anomaly detection technique is generic and can be extended to various types of ranging disturbances with negligible computational overhead.