Week 12: Choosing which channels to defend

The closing lecture — turning a semester of separately-argued channels into one design method with a stated stopping point

10 May 2027

No defender closes every channel, so the semester's output is a method for deciding which ones a stated device and adversary make worth closing.

Open the slides

No device is threatened by every channel this course has covered, and no defender has the budget to close every channel a device is theoretically exposed to. The semester’s last argument is that this is a design problem with a method, not a list to be worked through in order: state the device, state the adversary’s proximity and resources, use week 10’s taxonomy to rule out channels the adversary cannot reach or cannot afford, and rank what remains by the mutual-information estimate that quantifies how much each one actually reveals.

That method is exactly what the capstone design review asks a student to apply once, in full, to a device and adversary of their own choosing — the same six moves this lecture just walked through, run without a lecture supplying the answer. A course that spent twelve weeks establishing that channels are real ends by insisting that “real” is not the same question as “worth defending against, here, for this device” — the second question is the one an engineer is actually paid to answer.

Outline#

• the method, as six moves: state the device, state the adversary’s proximity and budget, rule out the channels that adversary cannot reach, rank what survives by rate against cost to close, name what you are choosing not to defend, and say what would change your mind
• the two failure modes in practice — an adversary stated so loosely that every channel applies, and a ranking with no residual risk admitted
• where the method stops: week 9’s placement defence, and the fact that the cheapest move is often not on the channel list at all
• the boundary week 6 drew, restated here as a limit on the method rather than on a week: this ranking covers channels that leak what a program did, and is silent on transient execution, which leaks what a program only speculated — a device exposed to both needs a second analysis this course has not taught, and a design review that does not say so has overclaimed its coverage

Readings

1. Qian Ge, Yuval Yarom, David Cock and Gernot Heiser (2018). A Survey of Microarchitectural Timing Attacks and Countermeasures on Contemporary Hardware. Journal of Cryptographic Engineering 8(1), 1-27.
2. Francois-Xavier Standaert, Tal G. Malkin and Moti Yung (2009). A Unified Framework for the Analysis of Side-Channel Key Recovery Attacks. EUROCRYPT 2009, LNCS 5479, 443-461.
3. Ross Anderson (2020). Security Engineering, third edition, chapter 19: Side Channels. Wiley.