Symmetric Cryptanalysis II
summary
Uses a SAT solver to evaluate differential clustering in multiple-branch designs and applies it to Orthros, raising the evaluated security bound by one round and achieving about 90% reduction in SAT invocations and 99% reduction in runtime versus the general method while producing higher differential probabilities. Introduces Rotational‑XOR Differential Rectangle (RXDR) cryptanalysis that mixes RX and classic differentials, finds optimal RXDR distinguishers up to 13–16 rounds for small SIMON variants and 15–20 for SIMECK, and reports experimental probabilities larger than theoretical values and a full key‑space distinguisher for SIMECK32 at 16 rounds.
AI-written summary · about
chapters (20)
- 0:00Introduction to research
- 2:24Background and motivation
- 4:38Multiple-branch designs challenge
- 7:14Proposed SAT-based strategy
- 8:43Five-step method explained
- 11:06Application to Orthros cipher
- 13:29Experimental results comparison
- 15:02Optimization with minimum weight
- 17:04Rotational-XOR rectangle cryptanalysis
- 18:07RX cryptanalysis basics
- 19:13Rectangle cryptanalysis overview
- 20:18RXDR cryptanalysis idea
- 23:31Differential rotational-invariant property
- 25:21Extending distinguishers with clustering
- 27:13Calculating multiple differentials
- 29:03Results for SIMON and SIMECK
- 30:35Experimental verification process
- 33:36Comparison with prior work
- 34:42Conclusion and future insights
- 35:31Q&A session
