GPS-Denied Navigation for Drones: Flying Without Satellite Positioning

Most drones lean heavily on GNSS — GPS and its equivalents — to know where they are. But satellite signals are weak and easily disrupted. In contested environments, jamming and spoofing are routine, so any serious autonomous platform needs to keep flying accurately when GPS is unavailable or untrustworthy. This is what "GPS-denied navigation" means and how it is done.

Why GPS fails in the field

GNSS signals arrive from orbit at very low power, which makes them fragile in two ways:

• Jamming — a stronger local transmitter drowns out the satellite signal, so the receiver simply loses its fix.
• Spoofing — a deceptive signal mimics real satellites and feeds the drone a false position, which is more dangerous than losing GPS entirely because the platform may not know it has been fooled.

Both are cheap and widely available, which is why relying on GPS alone is a single point of failure for any drone expected to operate in contested airspace.

How drones navigate without GPS

There is no single replacement for GPS; robust navigation comes from combining several complementary methods, none of which depends on an external signal:

Inertial navigation (INS/IMU)

Accelerometers and gyroscopes measure the platform's own motion, so it can dead-reckon its position. Inertial systems are self-contained and impossible to jam, but small errors accumulate ("drift") over time — so they are usually paired with an aid that periodically corrects the drift.

Visual and visual-inertial odometry

Cameras track features in the environment frame to frame to estimate motion. Combined with the IMU (visual-inertial odometry), this gives accurate short-to-medium range positioning without any external reference.

Terrain- and map-relative navigation

The platform matches what it sees below against known terrain or imagery to fix its absolute position — effectively recognising where it is from the landscape rather than from a satellite.

Sensor fusion is the real answer

The strength of a GPS-denied system is not any one sensor but how it fuses them. A fusion filter weighs each source by how trustworthy it is at that moment — leaning on vision in feature-rich areas, on inertial data through brief gaps — and can flag when a GPS fix looks like spoofing and should be ignored. The result is continuous, resilient positioning that degrades gracefully instead of failing outright.

A drone that stops working the moment GPS drops is not autonomous — it is remote-controlled by a signal an adversary can switch off.

This capability is foundational for autonomous swarms and is coordinated at the mission level by a C6ISTAR platform.