ECE 495 — Electromagnetic Warfare and Survivability

Block 2: RF Electronic Warfare
Lesson 18: Active Deception

Juan Jurado, Col, USAF, Ph.D.
Permanent Professor and Head
Electrical and Computer Engineering
Fall 2026

←→ navigate · C draw · B board · S speaker view · F fullscreen · ? all shortcuts

Block 2 · Lesson 18 · Where We Left Off

Where We Left Off

L17 was noise: raise the floor until the real return drowns. Brute force — and burn-through eventually beats it.

Deception is the surgical option. Instead of hiding the target, hand the radar a convincing lie and let it track that.

Today

DRFM, false targets, gate pull-off, and decoys — the deception toolkit.

Block 2 · Lesson 18 · Objectives

Lesson Objectives

By the end of this lesson, you will be able to:

Contrast deception with noise jamming
Explain how a DRFM repeater builds believable false returns
Walk through range-gate pull-off step by step
Describe towed and expendable decoys

Part 1

Win the Bin

A tracker keeps its books in cells. Noise buries a cell; deception fills it with something believable — and the radar tracks that instead.

Block 2 · Lesson 18 · Win the Bin

Deception: Make the Radar Track a Lie

Deception is EA — the branch that lies instead of shouting. A tracker keeps its books in bins: a range cell, a Doppler cell, an angle cell — one per quantity it follows.

Noise masks a bin; deception fills or shifts it with a false but believable return
The radar locks the strongest, most plausible return — not always the real one
Win the bin and you own what the radar thinks is true

Noise denies the picture; deception paints a fake one.

Block 2 · Lesson 18 · DRFM

From Transponder to DRFM

Transponder

Reply to each pulse after a fixed delay. Makes a range-false target, but non-coherent and easy to flag.

DRFM

Digital RF Memory: digitize, store, and replay a phase-coherent copy under software control.

DRFM sets precise delay, Doppler, amplitude, and can replicate pulses — false returns that look real

A coherent copy earns the radar’s own matched-filter gain — the lie is as loud as the truth.

Block 2 · Lesson 18 · False Targets

False-Target Injection

Fill the range and Doppler bins with believable replicas:

A search radar drowns in plausible contacts — which one is real?
A tracker can be seduced into locking the wrong return

On the scope every one of those spikes looks like this one. Only we can see which is which.

Part 2

Stealing the Gate

False targets buy confusion. To break a lock you have to reach into the gate itself — cover it, walk it out, and then disappear.

Block 2 · Lesson 18 · Range-Gate Pull-Off

Stealing the Gate: The Cover Pulse

t = 0 s · gate locked on the target

GATE · 150 m

range → offset from the true target (m)

Cover pulse floods the range bin — the gate locks the jammer, not the target
Grow the delay — the false return walks out and drags the gate with it
Blink off — the gate coasts in empty space; the target has escaped

Block 2 · Lesson 18 · Range-Gate Pull-Off

Stealing the Gate: Walk It Out

t = 2.5 s · gate at +750 m — skin return outside

GATE · 150 m

range → offset from the true target (m)

Cover pulse floods the range bin — the gate locks the jammer, not the target
Grow the delay — the false return walks out and drags the gate with it
Blink off — the gate coasts in empty space; the target has escaped

Block 2 · Lesson 18 · Range-Gate Pull-Off

Stealing the Gate: Blink Off

t = 5 s · jammer off — gate stranded at +1500 m

GATE · 150 m

range → offset from the true target (m)

Cover pulse floods the range bin — the gate locks the jammer, not the target
Grow the delay — the false return walks out and drags the gate with it
Blink off — the gate coasts in empty space; the target has escaped

VGPO does the same in Doppler; together they pull a range-Doppler diagonal.

Block 2 · Lesson 18 · Angle Deception

Angle Deception: Inverse Gain

Range and Doppler are not the only bins — the angle track can be fooled too.

A monopulse radar reads angle error from its $\Delta/\Sigma$ (difference-over-sum) ratio
Inverse-gain jamming transmits a phase-inverted $\Delta$ to null that ratio
The radar reads zero angle error — the tracker freezes or drifts off

Modern digital monopulse and ECCM blunt this — but the principle still drives decoy design.

Block 2 · Lesson 18 · Decoys

Decoys: Towed and Expendable

If you cannot fool the gate from on board, give the radar a better target to lock:

Towed decoy

A repeater trailed behind the aircraft; brighter and offset, it seduces the lock away.

Expendable decoys

MALD, ITALD — saturate the picture, bait radars into radiating, transfer the lock.

Active decoys beat chaff — no zero-Doppler notch, and they look real

The safest target to shoot at is the one that is not you.

Block 2 · Lesson 18 · Counters

Deception Isn’t Free

Every trick has a counter — the ECCM half of the duel:

Leading-edge tracking beats RGPO — the real skin return always arrives first
Digital monopulse and track logic blunt angle deception
Processing gain, polarization, waveform agility expose the copy
A coherent DRFM replay still leaves fingerprints — timing and phase artifacts

Deception buys time and doubt, rarely a permanent kill of the track.

Today’s Interactive Demo: DRFM Deception

Companion script: L18_RangeGatePullOff.m
In L19 the whole Block 2 toolkit goes to work — Project 2 work day

Live demo

DRFM Deception

usafa-ece.github.io/ece-495-ew

Demo hand-off; drive it from here. The scope is a range axis measured as offset from the true target, so zero is the skin return and the whole picture spans about two kilometres — the scale the lesson and the homework actually work in. The true target is painted navy at zero and holds a fixed thirty kilometre range; the DRFM false return is red; the tracking gate is the blue bracket, drawn to scale at one hundred fifty metres, which is L4's one microsecond pulse. Press Play and the sequence runs in three stages named in the banner: capture, pull-off, then blink and hold. In capture the gate is already locked on the skin return and the cover pulse grows louder on top of it until the gate follows the copy. In pull-off the false return walks out and drags the gate with it; the dashed red line at seventy-five metres is half a gate width, and when the gate passes it the skin return falls outside the gate and the readout flips to lock broken. At the default three hundred metres per second that happens a quarter second into the walk. The walk runs out to fifteen hundred metres over five seconds, then the jammer blinks off and the gate holds the last range it was dragged to. The separation callout reads in metres and in gate widths; the four cards are gate offset, the true target's range, whether the skin return is still in the gate, and the stage. The rate slider runs one hundred to one thousand metres per second — walk it to the top and ask what a tracker's rate logic would make of the fastest setting.

Block 2 · Lesson 18 · Quick Exercise

Quick Exercise / Type-along: When the Target Leaves the Gate

Type this with me — L4’s gate, and a DRFM walking at a believable closing speed.

c    = 3e8;                     % m/s
tau  = 1e-6;                    % radar pulse width, s (L4)
gate = c*tau/2;                 % range gate width, m
walk = 300;                     % DRFM range rate, m/s (believable)
t    = [0.1 0.3 1 2.5 5];       % time since cover pulse, s
off  = walk*t;                  % pull-off distance, m
dly  = 2*off/c;                 % added DRFM delay, s
out  = off > gate/2;            % true target outside the gate?
fprintf('gate = %.0f m wide\n', gate);
fprintf('  t(s)  pull(m)  delay(us)  out?\n');
fprintf('%6.2f %8.0f %10.2f %5d\n', [t; off; dly*1e6; out]);

The gate is 150 m deep. When does the true target leave it?

And that 1500 m of pull-off is only 10 µs of delay — why does leading-edge tracking still win?

Block 2 · Lesson 18 · Quick Problem

Quick Problem: Size the Pull-Off

A tracker with a shorter pulse than L4’s, and a DRFM that must walk its range gate off the target.

Pulse width $\tau = 0.4$ µs, so the range gate is $c\tau/2 = 60$ m deep
Blink off after five gate widths of pull-off: 300 m
The rate filter rejects any walk faster than 250 m/s
$c = 3\times10^8$ m/s

(a) How long does the walk take?
(b) What delay ramp must the DRFM program, in µs per second, and what is the total added delay at blink-off?
(c) The radar switches to leading-edge tracking. By how many µs does the false return trail the skin return at blink-off, and how many pulse widths is that?

(a) 1.20 s · (b) 1.67 µs per second; 2.00 µs total · (c) trails by 2.00 µs = five pulse widths

Two to three minutes, calculators or a MATLAB console; then click. Worked path: the gate is c tau over 2, 60 m, so five gate widths is 300 m. (a) 300 m at 250 m/s is 1.20 s — and the true target was already clear of the gate after half a gate width, 30 m, which is 0.12 s in; the rest of the walk parks the gate where reacquisition is slow. (b) Invert delta R equals c delta t over 2: delay is 2 delta R over c. The ramp is 2 times 250 over 3e8, 1.67 microseconds of delay per second of walk, and the total at blink-off is 2 times 300 over 3e8, 2.00 microseconds; check, 1.67 times 1.20 is 2.00. (c) A DRFM can only add delay, so at blink-off the false return trails the skin return by that same 2.00 microseconds, which is five 0.4-microsecond pulse widths. A leading-edge tracker gates on the first-arriving energy, the skin return, so the whole 1.2-second walk moves nothing. The so-what: the pull-off that fools a centroid tracker is trivially separable in time from the truth — that is why leading-edge tracking is the counter.

Block 2 · Lesson 18 · Wrap-Up

Wrap-Up

Key ideas

Deception fills or shifts the tracker’s range, Doppler, and angle bins
DRFM replays a coherent copy — a false return as loud as the real one
Cover pulse then pull-off walks the gate off the true target
Decoys give the radar a brighter target to lock; ECCM fights back

Next lesson

L19: Project 2 work day — B-21 RWR AoA characterization
Block 2’s EW toolkit now goes to work on a real scenario