Researchers Look to the Stars to Provide Drone Navigation When GPS Is Unavailable

When GPS is being jammed, a Raspberry Pi 5-powered modern sextant can provide a location fix for navigation.

Gareth Halfacree
2 months agoDrones / HW101

Researchers from the University of South Australia, working with the Defence Science and Technology Group, have come up with a solution for drone navigation in the absence of Global Navigation Satellite System (GNSS) availability — and it'd be familiar to seafarers of yore: location-finding via the stars.

"In environmental monitoring over remote locations or long-endurance surveillance missions where GPS might be unavailable or compromised, this technology offers a valuable new capability," says co-author Javaan Chahi of the team's work. "GNSS denial is an increasing challenge, and our research addresses this gap. We have developed a navigation method that’s resilient, independent of external signals, and achievable with low-cost, easily accessible components. This makes it applicable to a variety of UAVs [Uncrewed Aerial Vehicles], from commercial drones to more advanced defense applications."

The modern, automated version of a sailor's sextant could help autonomous drones navigate when GNSS is being jammed, researchers say. (📹: University of South Australia)

When GNSS systems like GPS aren't available — either due to failure, atmospheric conditions, or active jamming or spoofing attacks — drones and other uncrewed vehicles need a reliable way to navigate. The team's solution is to look skyward, charting a course by the stars like a sailor with a sextant.

"Unlike traditional star-based navigation systems, which are often complex, heavy and costly, our system is simpler, lighter and does not need stabilisation hardware, making it suitable for smaller drones," corresponding author Samuel Teague explains. "This type of navigation is ideal for operations over oceans, or in warfare zones where GPS jamming is a risk. Apart from the defence sector, it could also be highly useful for environmental monitoring."

The prototype navigation system is built using an ArduPilot Cube Orange flight controller linked to a Raspberry Pi 5 single-board computer and Alvium 1800 U-240 monochrome image sensor, and is capable of positioning a drone within 4km (around 2.5 miles) — a considerably less accurate fix than delivered by GNSS, but one that is available wherever you can see the stars — by flying in fixed orbits until a fix is delivered.

"Testing found that the algorithm is robust against initial conditions," the researchers say of field experiments using a fixed-wing drone, "requiring no knowledge of the prior position and only requiring the camera to be aligned to within a hemisphere of tolerance."

The stellar approach stands in exact opposition to that currently under test by NILEQ and Advanced Navigation to solve the same problem, in which a camera is trained on the Earth's surface rather than the stars — treating the undulating surface like a fingerprint in order to provide a GNSS-free location fix.

The team's work has been published in the journal Drones under open-access terms.

Gareth Halfacree
Freelance journalist, technical author, hacker, tinkerer, erstwhile sysadmin. For hire: freelance@halfacree.co.uk.
Latest articles
Sponsored articles
Related articles
Latest articles
Read more
Related articles