Reading — Triangulation and Fusion#
By the end of this lesson you should be able to:
Triangulate an emitter by crossing two lines of position into a fix.
Explain how crossing geometry, not receiver quality, sets the fix error.
Describe multi-receiver least-squares fusion and why more bearings help.
Trace the full RWR processing chain from intercept to track.
One bearing is a line, not a place#
L13 turned a measured phase difference into a single number: a bearing — the direction to the emitter. That number is genuinely useful, but it answers only half the question. A bearing tells you which way the threat is; it says nothing about how far. The emitter could be a fighter 20 km out or a ground site 200 km out, and the same bearing would point at both.
The honest way to draw a bearing is therefore not as an arrow but as a ray that runs off toward the horizon — a line of position (LOP). The emitter is somewhere on that line. To pin it to a point, you need a second line that crosses the first.
Two lines cross at a fix#
Put a second receiver somewhere else and have it measure its own bearing to the same emitter. Now you have two LOPs, drawn from two known locations. They intersect at exactly one point, and that point is the emitter’s position — a fix. This is triangulation: two angles measured from the two ends of a known baseline (the segment joining the receivers) determine the third corner of the triangle.
Nothing here is new geometry — it is the same construction a surveyor uses. What makes it EW is that the angles are noisy bearings squeezed out of a phase interferometer, and the “third corner” is a hostile radar that does not want to be found. Two receivers, two bearings, one point — if the geometry cooperates. That last clause is the whole lesson.
Key Concept
One bearing is a line of position: the emitter lies somewhere along it, near or far — you cannot tell which. Cross a second bearing from a second receiver and the lines meet at a fix. Triangulation converts two directions into one position by leaning on the known baseline between the receivers.
The cut angle sets the error#
Real bearings are not infinitely thin lines. Each one carries a noise band — a degree or two of angular uncertainty (L13) — so each LOP is really a thin wedge that fans out with range. The fix lives where the two wedges overlap, and the shape of that overlap is what we actually mean by “fix error.” The single thing that controls it is the crossing angle, or cut, at which the two bearings meet.
When the cut is near \(90^\circ\), the two wedges overlap in a small, roughly circular patch. Error in either bearing pushes the fix only a little, and it pushes in different directions, so the result is tight in every direction. This is the geometry you want.
When the cut is shallow — the two bearings nearly parallel — the wedges slide along each other and the overlap stretches into a long, thin sliver. A tiny bearing error now slews the intersection a long way down the line of sight. The fix is still decent across range but smears badly along it. This is exactly the case of an emitter far downrange from a short baseline: the receivers see it in almost the same direction, the cut collapses, and the error ellipse elongates toward and away from the receivers.
Push that to its limit and the emitter falls onto the line through the two receivers — a seam (collinear emitter and receivers). Now there is no cut at all: both receivers report the same bearing, the two LOPs lie on top of each other, and they share infinitely many “intersections.” Range along the baseline becomes unobservable — triangulation simply cannot tell you how far down the seam the emitter sits. Move even slightly off the seam and the fix returns, badly stretched, then tightens as the cut opens back toward \(90^\circ\).
Key Concept
Crossing geometry — not receiver quality — sets the fix error. A \(\sim 90^\circ\) cut gives a tight, near-circular fix; a shallow cut smears the error along range; a collinear seam makes range unobservable. The same receivers can deliver a precise fix or a useless one depending only on where the emitter sits relative to the baseline. (This is the geometric intuition behind GDOP; the formal information-theoretic treatment — Fisher information and the CRLB — waits for Block 4.)
More receivers: least-squares fusion#
With three or more receivers, the bearings almost never meet at a single point. Noise nudges each LOP a little, and three near-misses form a small triangle rather than a clean intersection. That is not a failure — it is information. You no longer have just enough equations to solve for the two unknowns (the emitter’s north/east coordinates); you have more equations than unknowns. The problem is overdetermined.
The fix is then the position that best fits all the bearings at once — the point that minimizes the total mismatch between where each receiver says the emitter is and where the candidate position would put it. That is a least-squares solution. Two payoffs follow:
Redundancy. Extra bearings average down the noise, and a well-placed receiver can open the cut that two co-located receivers could not. More geometry, better fix.
Outlier detection. A bad bearing — a reflected signal, a misassociated pulse — no longer hides. It shows up as a large residual: its LOP misses the consensus fix by far more than the others. You can spot it, down-weight it, or throw it out and re-solve.
Two bearings give you a fix. More bearings give you a fix you can trust — and a way to know when one of your own measurements has gone wrong.
Networked ESM: registration, association, fusion#
Crossing bearings gets powerful when the receivers sit on different platforms — wingmen, ships, ground sites — separated by tens or hundreds of kilometers. Each platform intercepts the emitter, builds pulse descriptor words (PDWs), measures its own bearing, and shares the result over a data link. Three steps turn that scattered data into a fix:
Registration. Every receiver must report on a common time and a common grid. A bearing measured at the wrong instant, or referenced to the wrong position, poisons the fix. Time and space registration come first.
Association. Decide which measurements belong to the same emitter across the receivers. (More on this below — it is the hard step.)
Fusion. Cross or least-squares the associated bearings into a position.
The payoff of a networked geometry is the wide baseline. Two receivers a few meters apart on one aircraft collapse to a shallow cut almost immediately with range. Two platforms 100 km apart hold a strong, near-\(90^\circ\) cut against an emitter hundreds of kilometers away. This is the B-21 standoff case: a formation can geolocate a threat radar from well outside its lethal range, fusing bearings to place an emitter that no single receiver could ever locate on its own. It is the bridge from warning (“there’s an SA-21 out there, roughly that way”) to targeting (“the SA-21 is here, to within a kilometer”).
The full RWR processing chain#
Everything in L12–L14 snaps into one pipeline that runs from the antenna to the cockpit:
Intercept → PDW → Deinterleave → AoA is the single-receiver top row from L12: catch a pulse, parameterize it into a PDW (frequency, PRI, PW, amplitude, time), sort the interleaved pulse trains back into individual emitters, and measure a per-pulse angle of arrival.
Associate → Fuse/geolocate → Track → Cue is the across-receiver bottom row that L13–L14 add: match emitters between receivers, cross their bearings into a located fix, track that fix over time into a moving threat, and cue it to the display, the missile-warning logic, or the strike package.
Read the chain and you can place every lesson in Block 2’s ES arc. L12 built the top row on one receiver; L13 turned phase into the bearing that feeds AoA; this lesson closes the bottom row, turning many bearings into a tracked, located threat.
Association comes before crossing#
There is a trap hiding in the word “fuse.” Crossing two bearings assumes the two receivers heard the same emitter. In a sparse sky that is safe. In a dense scene — a dozen radars, hundreds of overlapping pulse trains — it is a real question, and getting it wrong is worse than getting nothing.
So association precedes geolocation: before you cross anything, you match candidate emitters across receivers on their fingerprints — frequency, PRI, and pulse width — exactly the PDW parameters the deinterleaver already pulled out. Only once two reports agree that they describe the same emitter do you cross their bearings.
Cross the wrong pair — receiver A’s bearing to emitter X with receiver B’s bearing to emitter Y — and the two lines still intersect somewhere. That intersection is a ghost: a confident, precise-looking fix on an emitter that does not exist. Ghosts are the characteristic failure of multi-receiver geolocation, and they come not from bad bearings but from bad bookkeeping. Geolocation is only ever as good as the association feeding it.
Quick Exercise
Two receivers sit 8 km apart on an east–west line. Reason about each case:
An emitter sits due north, abeam the pair. Good cut or bad?
The same emitter is now 40 km north. What happens, and which way?
The emitter drifts onto the receivers’ east–west line. What breaks?
You may reposition one receiver. Where do you put it, and why?
Solution
Abeam → wide cut (~\(90^\circ\)) → a tight, near-circular fix. With the emitter directly north of the baseline’s midpoint, the two bearings splay apart and cross almost perpendicularly — the best geometry you can get.
Far downrange → shallow cut → the error smears along range (north). From 40 km out, both receivers see the emitter in nearly the same direction, so the cut collapses. Cross-range (east) stays reasonable, but the fix elongates toward and away from the receivers, and the overall error grows with range.
The seam. Collinear emitter and receivers means no cut at all: both LOPs lie on the east–west line, range along the baseline is unobservable, and the fix runs off to infinity along that direction.
Move one receiver to open the cut back toward \(90^\circ\) at the emitter — i.e., widen the baseline perpendicular to the line of sight. You are not buying a better receiver; you are buying better geometry, which is what actually sets the error.
Wrap-Up#
One bearing is a line of position; cross a second from a second receiver and the lines meet at a fix. The cut angle — not the quality of the receivers — sets the error: a \(\sim 90^\circ\) cut gives a tight, near-circular fix, a shallow cut smears it along range, and a collinear seam makes range unobservable. Adding receivers turns the fix into an overdetermined least-squares solution with redundancy and outlier-spotting built in. Networked across platforms, a wide baseline holds a strong cut at long range — the standoff geometry that lets a B-21 formation geolocate a threat it could never approach. The whole pipeline runs intercept → PDW → deinterleave → AoA → associate → fuse → track → cue, and the one step that cannot be skipped is association: cross the wrong pair and you conjure a ghost. Next, L15 — Electromagnetic Protect flips to the radar’s side of the duel: LPI waveforms, frequency agility, and side-lobe cancellation, all designed to deny the listener everything this lesson just exploited.