Drone localisation and navigation enhancement using radio wave sensing
File(s)
Author(s)
Awad, Heba
Type
Thesis or dissertation
Abstract
Drones have become essential in various applications, increasing the need for robust systems to ensure safe and efficient operation. This thesis investigates the use of radio wave sensing for drone detection, localisation, and state estimation, offering a cost-effective and resilient alternative to optical systems, which can be hampered by environmental factors like lighting and obstructions.
The thesis leverages radio wave technology to sense both the coarse- and fine-grained movements of drones, demonstrating their impact on drone localisation and navigation stability.
First, the thesis investigates the problem of sensing fine-grained drone movements during various flight situations. Two radio wave sensing technologies (5GHz RF OAM and 77GHZ mmWave radar) are adopted and evaluated to capture drone propeller speeds based on the Doppler shift effect. Then, the thesis proposes a passive indoor drone detection and 3D tracking system using a single-chip mmWave radar. This system combines extracted drone propeller rotation signatures with other mmWave radar measurements including range and angle, to achieve precise drone detection and motion type recognition, thereby enhancing the effectiveness of drone localisation. Finally, the thesis employs the radio-wave-sensed propeller speeds to enhance the autopilot’s state estimation and motion control of the drone.
Extensive experiments are conducted in different flying arenas using different drone models and sizes to execute a range of aerial manoeuvres, while encountering various line-of-sight and non-line-of-sight scenarios within dynamic environments to demonstrate the effectiveness of developed approaches. The experimental results highlight the system's ability to precisely discern the drone's coarse- and fine-grained movements and its robustness in detecting drones even in complex environments with moving objects, resulting in a significant reduction in localisation error compared to the classical baseline approach. Moreover, the system achieves enhanced drone trajectory tracking and improved navigation stability, even in challenging scenarios such as motor faults.
The thesis leverages radio wave technology to sense both the coarse- and fine-grained movements of drones, demonstrating their impact on drone localisation and navigation stability.
First, the thesis investigates the problem of sensing fine-grained drone movements during various flight situations. Two radio wave sensing technologies (5GHz RF OAM and 77GHZ mmWave radar) are adopted and evaluated to capture drone propeller speeds based on the Doppler shift effect. Then, the thesis proposes a passive indoor drone detection and 3D tracking system using a single-chip mmWave radar. This system combines extracted drone propeller rotation signatures with other mmWave radar measurements including range and angle, to achieve precise drone detection and motion type recognition, thereby enhancing the effectiveness of drone localisation. Finally, the thesis employs the radio-wave-sensed propeller speeds to enhance the autopilot’s state estimation and motion control of the drone.
Extensive experiments are conducted in different flying arenas using different drone models and sizes to execute a range of aerial manoeuvres, while encountering various line-of-sight and non-line-of-sight scenarios within dynamic environments to demonstrate the effectiveness of developed approaches. The experimental results highlight the system's ability to precisely discern the drone's coarse- and fine-grained movements and its robustness in detecting drones even in complex environments with moving objects, resulting in a significant reduction in localisation error compared to the classical baseline approach. Moreover, the system achieves enhanced drone trajectory tracking and improved navigation stability, even in challenging scenarios such as motor faults.
Version
Open Access
Date Issued
2024-05-31
Date Awarded
2025-01-01
Copyright Statement
Creative Commons Attribution NonCommercial Licence
License URL
Advisor
McCann, Julie
Publisher Department
Computing
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)