L37 Pre-flight

L37 Pre-flight#

Block 4: Aircraft Survivability and Mission Planning — Lesson 37: Modern Trends in EW Work this before L37. Pre-flights are never collected — in class, anyone may be cold-called to present any question at the board, and that recitation is the participation grade.

Assumed pre-class reading#

  • L2’s free-space path loss, \(L_\text{fs} = 20\log_{10} R_\text{km} + 20\log_{10} f_\text{GHz} + 92.45\); L8’s \(P_d\)/\(P_{fa}\) trade and the ROC; L12’s threat library; L16’s processing gain and \((\text{J/S})_\text{after} \approx (\text{J/S})_\text{in} - G_p\).

  • The Jupyter Book Reading page for L37 — the cognitive-EW loop (sense → learn → adapt), reading an emitter classifier’s confusion matrix as \(P_d\) and \(P_{fa}\), GPS below the noise and the jammer’s denial radius, INS drift on a denied leg, and uplink versus downlink jamming.

Quiz questions (5 items, ~5 minutes)#

Q1. (Multiple choice) A notional GPS receiver sees the L1 C/A signal (1575.42 MHz) at \(-130\) dBm and tolerates a J/S of 30 dB. A jammer with an ERP of 3 W toward it sits 100 km away in free space. The J/S at the receiver, and the verdict, are:

  • [ ] (a) −11.6 dB — the jammer sits far below the signal, so lock holds easily

  • [ ] (b) −1.6 dB — the jammer arrives a little weaker than the signal, so lock holds

  • [ ] (c) 28.4 dB — louder than the signal but under the threshold, so lock holds

  • [ ] (d) 32.3 dB — 2.3 dB past the threshold, so the receiver loses lock

Q2. (Multiple choice) A GPS jammer denies a circle of some radius around itself. Its ERP is then quadrupled. An aircraft crosses the new circle straight through its center at the same speed, on a notional INS that drifts at a constant rate. Compared with crossing the old circle, the denial radius and the INS error at the exit:

  • [ ] (a) Both grow by \(\sqrt{2}\), because range goes as the fourth root of power

  • [ ] (b) Both double, because a one-way path loses power as \(R^2\)

  • [ ] (c) Both quadruple, because J/S at a fixed range grows in step with the ERP

  • [ ] (d) The radius doubles, and the error grows fourfold with the squared time

Q3. (Multiple choice) A notional emitter classifier is tested on 50 intercepts of each true emitter type. Rows are the true type; columns are what the classifier said.

True

said EW

said ACQ

said TTR

EW

45

3

2

ACQ

4

40

6

TTR

2

3

45

Treat “TTR” — a tracking radar, the call the aircrew cares about most — as the detection. The classifier’s \(P_d\) and \(P_{fa}\) for TTR are:

  • [ ] (a) \(P_d = 0.85\) and \(P_{fa} = 0.15\)

  • [ ] (b) \(P_d = 0.90\) and \(P_{fa} = 0.08\)

  • [ ] (c) \(P_d = 0.90\) and \(P_{fa} = 0.10\)

  • [ ] (d) \(P_d = 0.87\) and \(P_{fa} = 0.13\)

Q4. (Multiple choice) An adversary wants to deny a military satellite-communications network across an entire theater with one ground jammer. The satellite is a simple bent-pipe relay. What must the jammer do?

  • [ ] (a) Jam the uplink from inside the satellite’s receive beam, so the relay carries the jam to every user

  • [ ] (b) Jam the downlink from any one site in the theater, since the downlink reaches every user at once

  • [ ] (c) Jam the uplink from anywhere in view of the satellite, since its antenna hears all of Earth equally

  • [ ] (d) Jam the uplink from beside the users, since uplink jamming only hurts terminals near the jammer

Q5. (Short answer) (i) Name the three steps of the cognitive-EW loop and say what each does. (ii) Explain why an RWR that classifies against a fixed threat library fails against an agile, software-defined radar. (iii) An ML emitter classifier trained on three emitter types meets a fourth type it has never seen. What does it report, and why is that dangerous?

Contents