Systematic Evaluation of Randomized Cache Designs against Cache Occupancy
summary
Compares five designs—CEASER, CEASER-S, MIRAGE, ScatterCache, SassCache—using a uniform benchmarking strategy across multiple replacement policies and shows performance differences (some randomized designs even beat the baseline under certain policies). Implements three attacks: a covert channel, process fingerprinting, and a T‑table AES key recovery via cache occupancy; the AES recovery succeeds and is the first full key recovery on a randomized cache using occupancy, with tests at 40%, 50%, 70% occupancy and guessing‑entropy measurements. Finds MIRAGE especially vulnerable and SassCache most resilient.
AI-written summary · about
chapters (11)
- 0:00Randomized cache designs intro
- 1:38Cache attacks background
- 2:41Cache occupancy explained
- 3:46Motivation for evaluation
- 5:50Benchmarking strategy
- 6:54Performance impact analysis
- 7:55Covert channel setup & results
- 9:43Process fingerprinting findings
- 11:12AES key recovery method
- 14:02Key recovery results
- 14:44Summary of findings
related talks
39mSecure Cache Designs: State of the Art and ...
13mExposing Cache Timing Side-Channel Leaks through Out-of-Order Symbolic Execution
44mCache Me If You Can: Messing with Web Caching
35mPage Cache Attacks: Microarchitectural Attacks on Flawless Hardware
17mQuantifying and Mitigating Cache Side Channel Leakage with Differential Set
23mMalicious Management Unit: Why Stopping Cache Attacks in Software is Harder Than You Think
27mCacheD: Identifying Cache-Based Timing Channels in Production Software
28mStrong and Efficient Cache Side-Channel Protection...
