ECE 495 — Electromagnetic Warfare and Survivability

Block 3: IR EW and Signature Management
Lesson 23: IR Seekers, IRST, and FLIR

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 3 · Lesson 23 · Where We Left Off

Where We Left Off

L22 dissected the target: five components, ranked by $I_{\text{band}} \propto \varepsilon \cdot A \cdot f_{\text{band}}(T) \cdot \sigma T^4$, with the winner set by band, aspect, and time of day.

Today we cross to the other side of the duel. Who is looking, and how?

Today

Three sensors exploit that signature: the seeker guides a missile, the IRST hunts like a passive radar, and the FLIR builds a picture for a human.

Block 3 · Lesson 23 · Objectives

Lesson Objectives

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

Distinguish seekers, IRST, and FLIR by job, field of view, and who reads the output
Explain how a reticle seeker turns a target image into a steering signal
Argue why imaging seekers changed the countermeasure game
Estimate passive detection range with $R_{\max} \propto \sqrt{J\,\tau/\text{NEI}}$ and contrast it with radar scaling

Every sensor today is passive. Nothing it does gives the target a warning — which is why the last objective is the one Block 3 keeps coming back to.

Block 3 · Lesson 23 · Part One

Who Is Looking, and How?

Three passive sensors, three jobs — and the job dictates the design, not the other way around.

Block 3 · Lesson 23 · Taxonomy

Three Jobs, Three Sensors

Seeker

Job — put a weapon on target
Instant. FOV — narrow
Angular res. — just enough to home
Output reads — guidance loop

IRST

Job — search and track
Instant. FOV — narrow (scans wide)
Angular res. — high (points)
Output reads — computer

FLIR

Job — image the scene
Instant. FOV — selectable
Angular res. — lower (scene)
Output reads — human

All three are passive: they collect the L22 signature and emit nothing
Same physics, different jobs — the job dictates the design, not the other way around

Block 3 · Lesson 23 · Seekers

Reticle Seekers: Homing Without a Picture

How do you track with one detector and no image? Spin a patterned disk in front of it:

The reticle chops the target's energy into an AM signal; the modulation reveals where the target sits off boresight
Guidance nulls the error; conical scan refines it with a nutating image and PWM
Cheap, simple, 1950s-proof — this is the seeker of the classic MANPADS
Weakness: it sees a point of energy, not a shape

To a reticle seeker, a flare is just a better target.

Block 3 · Lesson 23 · Seekers

Seeker Evolution: From Tail-Chase to All-Aspect

The seeker story is a band-and-aspect story — straight from L22:

First generation: uncooled SWIR detectors saw only hot metal — rear hemisphere, tail-chase shots
Cooled MWIR unlocked the plume ($\sim$4.3 µm): the missile went all-aspect
Modern seekers bias the aim point off the plume toward the airframe
Guidance rides proportional navigation: null the line-of-sight rate, not the tail

Block 3 · Lesson 23 · Seekers

Imaging Seekers: Why the Picture Changed Everything

Replace the reticle with a staring focal plane array and the seeker sees a scene:

The target is now a shape with edges and structure, not a blinking point
Seeker logic can weigh spectral, spatial, and temporal cues — does it look, glow, and move like an aircraft?
That is exactly the information a decoy struggles to fake
Rosette and pseudo-imaging scans paved the way; the FPA finished the job

Block 3 · Lesson 23 · IRST

IRST: The Passive Radar

An IR search and track system does a surveillance radar's job without transmitting:

Scans a wide field of regard, tracks autonomously, presents a radar-like display
Silent: the target's RWR hears nothing — there is no beam to detect
Precise angles, no portrait — the computer needs a track, not a picture

The catch

Angles come easy, range does not. Passive ranging needs the target's own kinematics over time, or a wingman's geometry from a second baseline.

A radar display with no transmitter behind it — the first indication of trouble may be a missile in flight.

Block 3 · Lesson 23 · The Range Rule

How Far Can a Passive Sensor See?

The sensor triggers when the target's irradiance beats its noise floor:

Passive IR detection range

$$\frac{J \cdot \tau(R)}{R^{2}} \;\ge\; \text{NEI} \qquad\Longrightarrow\qquad R_{\max} \;\propto\; \sqrt{\frac{J\,\tau}{\text{NEI}}}$$

$J$ — the target's in-band intensity: exactly what L22 taught you to rank
$\tau$ — atmospheric transmittance: the tax collector from L21, worse with range
NEI — the sensor's noise floor: cooling and detector material set it
Radar scales as $R_{\max} \propto \sigma^{1/4}$: stealth starves the radar, not the plume

Stealth moves the radar duel into the IRST's backyard.

Block 3 · Lesson 23 · FLIR

FLIR: The Picture for the Human

Forward-looking infrared optimizes for interpretation, not homing:

Detailed imagery, selectable field of view — wide to find, narrow to identify
Operator tools shape the image: gain and level, white-hot / black-hot polarity
A video tracker closes the loop: centroid and correlation logic hold the aimpoint
Navigation, targeting pods, reconnaissance, search and rescue — one family, many missions

Narrow FOV is how you identify — and it is also the soda straw you are looking through while you do it.

Block 3 · Lesson 23 · Looking Ahead

Where This Is Headed: Fighting the Seeker

Every sensor today defined a vulnerability — exploiting them is Lesson 24:

Flares seduce a reticle seeker: hotter point wins the modulation contest
Imaging seekers shrug flares off — so countermeasures must get smarter too
DIRCM puts modulated laser energy into the seeker to break its lock
Suppression attacks $J$ itself — shrink the intensity, shrink every $R_{\max}$ in the sky

For the B-21 the goal is simpler: never give the seeker enough $J$ to start the duel.

Block 3 · Lesson 23 · Quick Exercise

Quick Exercise / Type-along: Where the Stern Advantage Went

Type this with me — L22's two aspects, one passive sensor.

R = linspace(0.5,80,1591);        % range grid [km]
a = 0.15; NEI = 1e-10;            % 1/km, W/m^2
E = @(J) J.*exp(-a*R)./(R*1e3).^2;   % irradiance [W/m^2]
Rmax = @(J) R(find(E(J) >= NEI,1,'last'));
Rs = Rmax(1000);                  % stern: plume + hot parts
Rb = Rmax(50);                    % beam: skin
fprintf('stern %.2f km   beam %.2f km\n', Rs, Rb);
fprintf('vacuum %.2fx   atmosphere %.2fx\n', ...
        sqrt(1000/50), Rs/Rb);
% Same sqrt(J) rule, no atmosphere: MANPADS 8 km at 300 W/sr
Jnew = 120;
fprintf('suppressed to %d W/sr -> %.2f km\n', ...
        Jnew, 8*sqrt(Jnew/300));

Two ratios print — where did the difference between them go?

Then: why may the last line drop $\tau$, when the first two cannot?

Block 3 · Lesson 23 · Quick Problem

Quick Problem: Seen or Not at 30 km

An IRST holds a notional aircraft at a fixed 30 km and looks at it twice: stern-on, then beam-on.

Range $R = 30$ km; extinction $\alpha = 0.1$ /km, so $\tau = e^{-\alpha R}$
Sensor noise floor NEI $= 2\times10^{-9}$ W/m²
Stern intensity $J = 900$ W/sr
Beam intensity $J = 25$ W/sr

(a) The stern irradiance at 30 km, in multiples of NEI. Detected?
(b) The same for the beam.
(c) The largest $J$ that stays undetected at 30 km.

(a) $4.98\times10^{-8}$ W/m² = 24.9 NEI, detected · (b) $1.38\times10^{-9}$ W/m² = 0.69 NEI, not detected · (c) 36.2 W/sr

Two to three minutes, calculators or a MATLAB console; then click. Worked path: tau is e to the minus 0.1 times 30, e to the minus 3, 0.0498; R is 3e4 meters, so R squared is 9e8 — meters, not kilometers, in the R-squared term, the units trap the reading flags. (a) E is J tau over R squared: 900 times 0.0498 over 9e8 is 4.98e-8 W/m^2, which over the 2e-9 floor is 24.9 NEI — detected, with room to spare. (b) Same path with J = 25: 1.38e-9 W/m^2, 0.69 NEI — below the floor, not detected. At a fixed range E goes as J, so (b) is just (a) scaled by 25 over 900. (c) Set E equal to NEI and solve for J: NEI times R squared over tau, 2e-9 times 9e8 over 0.0498, is 36.2 W/sr; the beam's 25 sits under the line. The so-what: to vanish at 30 km the stern must drop from 900 to 36 W/sr, a 25-fold cut, not a trim — and no RCS treatment touches a single factor in that arithmetic.

Today's Interactive Demo: IR Detection Range

Companion script: L23_IRDetectionRange.m

Live demo

IR Detection Range

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

Demo hand-off; drive it from here. It gives the target two L22-style intensities — stern, plume plus hot parts, and beam, skin only — applies the atmospheric loss and plots irradiance at the sensor against range. Where each curve crosses the sensor's NEI floor is that aspect's maximum detection range, marked with a dot. Open on the two-target preset, then put the question to them: compare what the vacuum square-root rule promises against what the readout actually gives, and ask where the stern advantage went. The intensity slider drives the stern curve only; the beam target is fixed, so the two stay comparable. Then walk the extinction and NEI sliders — haze, and a better-cooled detector. Say the forward pointer out loud at the end: in L24 we fight back, with flares, DIRCM and plume suppression. Companion script runs on their own.

Block 3 · Lesson 23 · Wrap-Up

Wrap-Up

Key ideas

Three jobs, three passive sensors: seeker homes, IRST hunts silently, FLIR shows a human
Reticle seekers track a chopped point of energy; imaging seekers track a shape — and shapes are hard to fake
Seeker history = detector history: cooled MWIR turned tail-chase into all-aspect
$R_{\max} \propto \sqrt{J\,\tau/\text{NEI}}$ — and unlike RCS, the plume ignores your stealth budget

Next lesson

L24: IRCM — flares, DIRCM, and plume suppression against everything we met today
Work the L24 pre-flight before class — you could be the one up front