USAFA DFEC

ECE 495 — Electromagnetic Warfare and Survivability

Block 4: Aircraft Survivability and Mission Planning
Lesson 37: Modern Trends in EW
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 4 · Lesson 37 · Where We Left Off

Where We Left Off

L31–L36 built a route against a known threat: fixed sites, library parameters, a GPS-true position, and links that always work.

Every one of those is an assumption. The adversary is working to break each of them.

Today: four trends already changing the fight — at survey level, with one number you can compute.

Block 4 · Lesson 37 · Objectives

Lesson Objectives

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

  • Describe the cognitive-EW loop and why a fixed threat library fails
  • Read an ML emitter classifier's confusion matrix as $P_d$ and $P_{fa}$ (L8)
  • Compute a GPS denial radius and the INS drift over a denied leg
  • Outline space-EW geometry: uplink versus downlink jamming
The next fight adapts faster than the last one's playbook.

Block 4 · Lesson 37 · Four Trends

Four Trends, Four Broken Assumptions

Each trend attacks something your Final Project quietly takes for granted.

TrendThe assumption it breaks
Cognitive EWthe threat library is complete, and the threat behaves as catalogued
AI/ML emitter IDa classifier's label is a fact, not a probability
NAVWARyou always know where you are — GPS is truth
Space EWthe satellites and links behind your support package are always there

Survey level today. The one computation is the GPS denial radius.

Block 4 · Lesson 37 · Cognitive EW

Cognitive EW: Sense, Learn, Adapt

SenseES · characterize
→
Learnmodel its behavior
→
Adaptrespond · score it
  • L12's RWR matches a library that ELINT built; the loop learns in the mission
  • The score feeds back: did the jamming change what the radar does?
  • It runs on both sides — a cognitive radar adapts its waveform to your jamming
What does the loop have that a library lookup does not?

Block 4 · Lesson 37 · Cognitive EW

Why a Fixed Library Fails

Threat library (L12)Cognitive loop
Knowscatalogued emitters and modesbehavior it observes now
Meets a new modeno match — or the wrong onecharacterizes it by what it does
Update cyclecollect, analyze, reprogram: weekswithin the engagement
Weak spotagile, software-defined emitterstrust: can you verify what it learned?

L12 already hinted at this: one PRI per emitter stopped matching the moment the TTR changed mode.

Block 4 · Lesson 37 · AI/ML Emitter ID

ML Emitter ID: Read the Confusion Matrix

A notional classifier, tested on 100 intercepts of each true emitter type (L7's EW, ACQ and TTR radars).

Truesaid EWsaid ACQsaid TTR
EW9262
ACQ9838
TTR11188
Read the TTR row
$$P_d = \frac{88}{100} = 0.88$$
Read the TTR column
$$P_{fa} = \frac{2 + 8}{100 + 100} = 0.05$$

Rows: what the emitter really was. Columns: what the classifier said. The diagonal is right.

Block 4 · Lesson 37 · AI/ML Emitter ID

One Matrix Is One Point on a ROC

The L8 trade, again
  • Lower the bar to call TTR: $P_d$ and $P_{fa}$ both rise
  • A missed TTR costs the aircraft more than a false one costs the crew — so where does the bar go?
Where it breaks
  • Out-of-library: an unseen mode still gets a known label
  • Adversarial: an emitter shaped to be mislabeled
The matrix has no unknown column. What does it say about an emitter it has never seen?

Block 4 · Lesson 37 · NAVWAR

GPS: a Signal Below the Noise

1
L1 C/A from 20,000 km arrives weak (notional, textbook)
$f = 1575.42\ \text{MHz}, \quad S \approx -130\ \text{dBm}$
2
Thermal noise in its 2 MHz sits ~20 dB above it (L8)
$-174 + 10\log_{10}(2.046\times10^{6}) \approx -111\ \text{dBm}$
3
1.023 Mchip/s of code over 50 bit/s of data: L16's processing gain
$G_p = 10\log_{10}\dfrac{1.023\times10^{6}}{50} \approx 43\ \text{dB}$
4
The loops need a margin $M$ after despreading (L16)
$(J/S)_{\text{tol}} = G_p - M$

Block 4 · Lesson 37 · NAVWAR

The Jammer's Reach: a Denial Circle

1
Jammer power at the receiver: ERP less the one-way path
$J = \text{ERP} - L_{\text{fs}}(R)$
2
Block 1's free-space path loss, in km and GHz (L2)
$L_{\text{fs}} = 20\log_{10} R + 20\log_{10} f + 92.45$
3
Denied wherever J/S passes $(J/S)_{\text{tol}}$: a circle around the jammer
$L_{\text{fs}}(R_{\text{deny}}) = \text{ERP} - S - (J/S)_{\text{tol}}$
The jammer does not have to out-shout the satellite's transmitter — only the satellite's 20,000 km of path.

Block 4 · Lesson 37 · Quick Exercise

Quick Exercise / Type-along: Draw the Denial Circle

Type this with me — one receiver, three jammers.

% L37 -- how far does a GPS jammer reach?
f   = 1.57542;               % GPS L1 [GHz]
S   = -130;                  % received C/A power [dBm]
Gp  = 10*log10(1.023e6/50);  % chips per data bit -> G_p (L16)
M   = 13;                    % margin the loops need after despread
JSt = Gp - M;                % tolerable J/S at the antenna [dB]
PW  = [1 4 16];              % jammer ERP toward us [W]
L   = 10*log10(PW*1e3) - S - JSt;          % path loss at J/S = JSt
R   = 10.^((L - 20*log10(f) - 92.45)/20);  % invert L2's FSPL [km]
fprintf('G_p = %.1f dB   tolerable J/S = %.1f dB\n', Gp, JSt);
fprintf('%4.0f W -> GPS denied inside %5.1f km\n', [PW; R]);

Four times the power, twice the radius. Why only twice?

Then: what would 6 dB more J/S tolerance do to the 16 W circle?

Block 4 · Lesson 37 · Quick Problem

Quick Problem: Inside the Circle?

A notional 2 W GPS jammer on a ridge; your receiver is 50 km from it.

  • GPS L1: 1575.42 MHz; received C/A power $-130$ dBm
  • Tolerable J/S: 30 dB
  • Jammer ERP toward you: 2 W
  • Range to the jammer: 50 km
  • $L_{\text{fs}} = 20\log_{10} R_{\text{km}} + 20\log_{10} f_{\text{GHz}} + 92.45$
  1. (a) The path loss, and the jammer power $J$ at the receiver.
  2. (b) The J/S, and whether the receiver keeps lock.
  3. (c) The denial radius of this jammer.
(a) 130.4 dB; $J = -97.4$ dBm  ·  (b) 32.6 dB, 2.6 dB over: lock lost  ·  (c) 67.7 km

Block 4 · Lesson 37 · NAVWAR

Jamming Denies; Spoofing Deceives

Jamming · denial
  • Raises the noise until the loops lose lock
  • The receiver knows: lost lock is a loud, honest failure
  • Answers: nulling antennas (L6, L15), more $G_p$, coast on the INS
Spoofing · deception
  • Counterfeit signals capture the loops, then walk the solution off
  • The receiver may not know: a confident, wrong position
  • Answers: authentication, INS cross-checks, integrity monitoring
Spoofing is L18's range-gate pull-off, aimed at your position. Which would you rather meet on a long denied leg?

Block 4 · Lesson 37 · NAVWAR

The Denied Leg: INS Drift Meets the Route

MEZ GPS denied J/S above tolerable denial radius jammer route GPS back planned clearance INS error grows while GPS is denied
Error at the exit
$$e \approx \dot{e}\,t, \qquad t = \frac{\ell}{V}$$
  • $\dot{e}$: a navigation-grade INS drifts ~1.5 km/h (notional); lower grades far faster
  • $\ell$: the route's chord through the circle, flown at speed $V$
  • A clearance is real only if it exceeds $e$
Final Project: a seam you thread 5 km from a MEZ is only a seam if you know where you are.

Block 4 · Lesson 37 · Space EW

Space EW: Uplink versus Downlink

receive beam transponder relays what it hears downlink to users uplink jammer outside the beam uplink jammer inside the beam downlink jammer local only
  • Uplink: a jammer inside the satellite's receive beam reaches every user
  • Downlink: the satellite is faint, so a small local jammer wins — near itself only
  • GPS denial is downlink jamming: that is why it is a circle
  • EP answers: spot beams, nulling antennas and onboard processing

Block 4 · Lesson 37 · Synthesis

How Four Blocks Set Up These Trends

BlockWhat it gave you for today
B1 · Radar fundamentalsFSPL for the jammer's reach; the ROC for reading a classifier
B2 · RF EWES classification, processing gain, J/S — the core of all four trends
B3 · IR EW and signaturesmulti-band fusion and EMCON — what an ML fuser and a cognitive ES hunt
B4 · Survivabilitya route whose margins must now absorb navigation error

None of today's trends needs new physics. Each one changes who adapts faster.

Block 4 · Lesson 37 · Final Project

Your Limitations Section: Pick One Trend

TrendWhat it could change in your route
NAVWARclearances past each MEZ must exceed the INS error on a denied leg
Cognitive EWlibrary-based detection ranges may be wrong: widen the margins you trust
AI/ML emitter IDfaster, surer identification raises $P_{I|T}$ and each site's hazard
Space EWSOJ and SEAD timing ride on links that may be denied: plan for late support
A limitation earns its place when it changes a number or a line on your map — say which, and by how much if you can.

Block 4 · Lesson 37 · Wrap-Up

Wrap-Up

Key ideas
  • Cognitive EW senses, learns and adapts in the mission — faster than any fixed threat library
  • A confusion matrix row is $P_d$ and a column is $P_{fa}$: each matrix is one point on L8's ROC
  • GPS is denied inside a circle set by ERP, FSPL and $G_p - M$; the INS drifts inside it
  • Uplink jamming must reach the satellite's beam and hits every user; downlink jamming is local
Next lesson
  • L38: Final Project Work Day — the IEEE template, source vetting, your figures
  • Work the L38 pre-flight before class
  • Draft your Limitations paragraph with one trend

Companion script: L37_GPSDenial.m — J/S versus range, denial radius versus power, INS error over a denied leg