Trace Analysis Lab

Reading an unannotated power trace against the program that produced it, without a lecture walking you through it first

Due: 19 March 2027 · Weight: 30%

The brief#

You are given one power trace and the source of the program that produced it. Say which operation the secret controls, and where in the trace that control is visible.

Week 3 walked through one trace as a class, slowly, with the algorithm alongside it. This lab supplies a different program — still simple, still a single secret-dependent branch somewhere in its execution — and asks you to do the same reading alone. The trace will not be as clean as the deck’s; part of what is being assessed is whether you can still make the call under real noise, not just under the deck’s tidied example.

What you submit#

An annotated image of the trace (the same style of annotation practised in the Bench) and a short written report: what operation you believe is secret-dependent, where in the trace you can point to that dependence, and one sentence on how you would check the finding against a second trace from the same program. No exploit code or working key-recovery script is required or wanted — the deliverable is the reading, not a tool.