Skip to content
Blog

What is terrain-referenced navigation (TRN)?

Published:

Terrain-referenced navigation (TRN) estimates an aircraft's position by comparing the shape of the ground below it with a stored digital elevation model. Unlike inertial or visual odometry, it gives an absolute position fix, which makes it a natural partner for methods whose error grows over time.

The idea behind TRN

As an aircraft flies, the height of the terrain beneath it rises and falls. That sequence of heights forms a profile that is often unique to a particular path over the ground. If the aircraft can measure this profile and has a map of terrain elevation, it can search the map for the place where the profile fits best.

The match gives a position in map coordinates, independent of how long the aircraft has been flying. That is what makes TRN an absolute method, in contrast to relative methods that accumulate drift.

What TRN needs

A TRN system combines a few ingredients:

  • A height-above-ground measurement, typically from a radar or laser altimeter.
  • The aircraft's own altitude, usually from a barometric altimeter combined with the inertial solution.
  • A digital elevation model (DEM) of the area, stored on board.
  • A short-term motion estimate, normally from inertial navigation, to know how the measured profile is laid out along the path.

How the matching works

Subtracting the measured height above ground from the aircraft's altitude gives the terrain elevation at each point along the path. The algorithm then compares this measured profile with candidate profiles from the elevation model.

Some systems correlate a batch of measurements at once; others run a recursive filter, such as a particle filter or a Kalman-type filter, that updates the position estimate with each new measurement. In both cases the result is fused with inertial navigation so the aircraft has a continuous, corrected position.

Strengths and limitations

TRN works without light and without any outside signal, and its accuracy does not degrade with flight time. Its performance, however, depends heavily on the terrain:

  • Over rugged, varied terrain the profile is distinctive and the match is sharp.
  • Over flat ground or water the profile carries little information and the match becomes ambiguous.
  • Repetitive terrain can produce several similar candidate positions.
  • Accuracy is limited by the resolution and quality of the elevation model and the altimeter.
GPS-denied navigation: methods compared →

Frequently asked questions

They are related but different. TRN matches terrain height profiles against an elevation model. Scene matching compares camera images with reference imagery. Both provide an absolute position and both depend on how distinctive the area is.

Not well. Over water or very flat ground the measured height profile has almost no variation, so there is little to match. TRN is normally combined with other methods to cover such areas.

Inertial navigation provides smooth, high-rate motion but drifts. TRN provides occasional absolute fixes but needs a motion estimate to lay out the measured profile. Fusing them gives a continuous position with bounded error.

A digital elevation model (DEM) is a grid of terrain heights covering an area. TRN uses it as the reference map against which measured ground heights are compared.