|
PQ-WB-KEM: White-Box ML-KEM-768 1.0
PQ-WB-KEM: Toward a White-Box Construction of ML-KEM-768 with Arithmetic Masking for M2M Communications - Mathematics (MDPI)
|
Differential Computation Analysis simulation.
Collaboration diagram for DCA Simulation:Differential Computation Analysis simulation.
Simulates DCA attacks to verify the effectiveness of arithmetic masking. Measures correlation coefficients between intermediate values and key hypotheses.
The Hamming weight leakage model is used:
\[ L(x) = HW(x) + \mathcal{N}(0, \sigma^2) \]
where HW(x) counts the number of 1-bits in the 12-bit representation of x mod q.
rho_max is the correct-key correlation; the maximum over the wrong-key hypotheses is 0.031, itself within the null noise band.
| Implementation | Traces | rho_max | Attack |
|---|---|---|---|
| Unmasked | 1,000 | 0.866 | Success |
| Unmasked | 10,000 | 0.853 | Success |
| 3-Share Masked | 1,000 | 0.031 | Fail |
| 3-Share Masked | 5,000 | 0.005 | Fail |
| 3-Share Masked | 10,000 | 0.011 | Fail |
The masked correct-key correlations are non-monotone because all three are sampling-noise-floor values (~1/sqrt(N)) rather than a decaying signal.
The operative criterion is that the correct-key correlation rise above the wrong-key spread. At 10,000 traces it does not: the correct-key rho_max = 0.011 sits below the 0.031 wrong-key maximum, about 1.1x the 1/sqrt(N) = 0.01 sampling-noise floor, so no first-order leakage is detectable and the attack fails. The boolean resistant field of dca_simulation.json applies the stricter naive rho_max < 0.01 floor test, which 0.011 marginally exceeds; the wrong-key comparison is what determines the outcome. These figures characterize the masked share intermediates under a gray-box Hamming-weight model, not the whole program: the clear product coordinate formed before masking still leaks at the unmasked ~0.85 level.