[Paper Review] Accurate 3D Localization for MAV Swarms by UWB and IMU Fusion

Jiaxin Li, Yingcai Bi+4 more|arXiv (Cornell University)|Jul 28, 2018

Indoor and Outdoor Localization Technologies14 references10 citations

TL;DR

This paper proposes an Extended Kalman Filter (EKF)-based fusion algorithm that combines Ultra-Wideband (UWB) ranging and Inertial Measurement Unit (IMU) data to enable 80Hz, highly accurate 3D localization for micro aerial vehicle (MAV) swarms in GPS-denied environments. The method achieves sub-meter accuracy with near-zero latency, significantly outperforming vanilla EKF in both accuracy and responsiveness, as validated in indoor VICON and real-world exhibition hall tests with a 6-MAV swarm.

ABSTRACT

Driven by applications like Micro Aerial Vehicles (MAVs), driver-less cars, etc, localization solution has become an active research topic in the past decade. In recent years, Ultra Wideband (UWB) emerged as a promising technology because of its impressive performance in both indoor and outdoor positioning. But algorithms relying only on UWB sensor usually result in high latency and low bandwidth, which is undesirable in some situations such as controlling a MAV. To alleviate this problem, an Extended Kalman Filter (EKF) based algorithm is proposed to fuse the Inertial Measurement Unit (IMU) and UWB, which achieved 80Hz 3D localization with significantly improved accuracy and almost no delay. To verify the effectiveness and reliability of the proposed approach, a swarm of 6 MAVs is set up to perform a light show in an indoor exhibition hall. Video and source codes are available at https://github.com/lijx10/uwb-localization

Motivation & Objective

• To address the limitations of standalone UWB and IMU systems in MAV swarm localization, particularly high latency and low bandwidth in pure UWB solutions.
• To develop a real-time, high-update-rate localization system suitable for controlling fast-moving MAV swarms in indoor, GPS-denied environments.
• To improve localization accuracy and stability by fusing UWB range measurements with IMU inertial data using an EKF framework.
• To validate the system’s performance in both controlled (VICON) and real-world (Changi Exhibition Centre) environments with multi-MAV operations.

Proposed method

• An Extended Kalman Filter (EKF) is used to jointly estimate MAV position, velocity, and IMU bias using UWB range measurements and IMU acceleration/gyroscope data.
• UWB two-way ranging with TDMA/CDMA protocols enables precise, synchronized distance measurements between MAVs (mobiles) and fixed anchors.
• The system employs innovation thresholding to reject outlier UWB measurements caused by multipath or signal blockage in dense environments.
• The EKF state vector includes 3D position, 3D velocity, and IMU bias states to model and correct for sensor drift and bias.
• A coordinate system is established from precise anchor placement using North-West-Up (NWU) coordinates, with anchors positioned to form a convex hull around the operation area.
• The algorithm runs at 80Hz, enabling real-time control with minimal latency, crucial for swarm coordination.

Experimental results

Research questions

• RQ1 Can UWB-IMU fusion via EKF achieve real-time 80Hz 3D localization for MAV swarms with minimal latency?
• RQ2 How does the proposed EKF fusion method compare to vanilla EKF in terms of accuracy and delay in indoor environments?
• RQ3 To what extent can the fusion algorithm maintain localization accuracy under signal occlusion and multipath interference in real-world indoor settings?
• RQ4 Can the system support stable, coordinated flight of multiple MAVs in a large-scale indoor environment without GPS or VICON?

Key findings

• The fusion EKF achieved a maximum positioning error of 0.39 m and a mean error of 0.16 m in VICON testing, significantly outperforming the vanilla EKF’s 0.71 m and 0.30 m, respectively.
• The fusion EKF exhibited near-zero latency, while the vanilla EKF showed a 500ms delay between estimated position and VICON ground truth, rendering it unsuitable for real-time control.
• In the Changi Exhibition Centre performance, the system enabled stable hovering and coordinated flight of a 6-MAV swarm, with clean localization trajectories observed in video evidence.
• The system effectively rejected erroneous UWB measurements through innovation thresholding, maintaining position estimates even during signal blockages from other MAVs.
• The estimated IMU bias showed reasonable scale and trend, suggesting the EKF correctly modeled and compensated for sensor bias during flight.
• Despite signal interference from multiple UWB devices and WiFi, the fusion algorithm maintained robust performance in a complex electromagnetic environment.