Understanding GDOP with Two Fingers — The Geometry of Good Anchor Placement

In the previous installment, we looked at the three reasons location tracking gets shaky in the field, and noted that accuracy differs from zone to zone even with no obstacles at all. Today we take on the reason behind that — GDOP — with no equations, just two pointing fingers.

Ⅰ. A Position Is Where Lines Meet

There is an old technique for finding your position on a hiking trail with a map and a compass. Pick two peaks visible in the distance and draw a line toward each on the map. Where the two lines cross is where you are. Location tracking works the same way at its core: each distance or time difference measured by an anchor forms a line (strictly speaking, a surface), and the tag sits where those lines meet.

But there is a trap here. The angle at which the lines cross determines how sharp the intersection is.

Ⅱ. The Two-Finger Experiment

Here is a thought experiment you can try right now. Two people stand at opposite ends of a room and point at the same object.

• When they point from clearly different directions (say, 90 degrees apart), anyone can see exactly where the two sight lines meet.
• When they stand almost in the same spot, the two sight lines become nearly parallel, and the intersection stretches into a vague "somewhere around here."

Even if the pointers (the anchors) are equally skilled, the diversity of directions they stand in decides how sharp the answer is. GDOP (Geometric Dilution of Precision) is the single number that expresses this degree of blurring.

Position error = GDOP × measurement error

When GDOP is close to 1, measurement quality carries over to position quality almost as-is. When GDOP is 5, the same measurement quality produces five times the position error. No matter how clean the signal, bad geometry can never yield good coordinates.

For the record, this concept came from GPS. It is the same principle behind your navigation drifting when the satellites, however numerous, cluster in one part of the sky. Indoor positioning faces a tighter constraint than GPS — you can rarely mount anchors exactly where you want — which makes GDOP matter even more indoors.

Ⅲ. The Error Map of a Real Site — Four Patterns

Inside an actual building, GDOP shows up in four patterns.

• The center of a large hall — anchors are visible evenly in every direction. GDOP is at its minimum; this is the most accurate zone.
• A long corridor — anchors line up in a single row, so positions are accurate along the corridor but weak across it. This is why dots jitter across the corridor's width.
• Spaces with ceiling-only anchors — horizontal position holds, but height (floor-level) estimates become unstable.
• Corners and alcoves — the visible anchors bunch to one side and GDOP spikes. This is where error spreads the widest.

One important fact here: directional diversity comes before anchor count. Four anchors spread in all directions often beat six anchors bunched on one side. That is why you should look at directions before adding anchors.

Ⅳ. When the Same Number Means Different Things

Let's go one step deeper. Even systems that do use GDOP often apply a single global threshold — "discard anything above GDOP 2." But a GDOP of 1.5 at the center of a hall and a GDOP of 1.5 in a corner mean different things. At the center, 1.5 is worse than usual; in the corner, 1.5 may be the best that space can ever offer. It is like body temperature: everyone's baseline differs, so 36.8°C is normal for one person and a slight fever for another.

That is why ORBRO divides a space into zones and measures each zone's usual GDOP baseline in advance. Judgments are then made relative to that zone's baseline, not against an absolute value. This is the starting point of the approach ORBRO's technical white paper calls "spatially adaptive."

In Closing

Today's takeaway fits in three lines: a position emerges where lines meet, the crossing angle determines its sharpness, and GDOP is the number that measures it. A good positioning system reads this number zone by zone and responds differently to each.

One question remains, though. If receiving more anchors is an advantage, why does the industry have so many systems that deliberately reduce signals? We tackle that paradox in the next installment.

If you are wondering what anchor layout fits your site's structure, contact ORBRO.