Analysis of Nonparametric Estimation Methods for Mutual Information Analysis - Information Security and Cryptology-ICISC 2010 - page 14

28. Venelli, A.: Ecient Entropy Estimation for Mutual Information Analysis Us-
ing B-Splines. In: Samarati, P., Tunstall, M., Posegga, J., Markantonakis, K.,
Sauveron, D. (eds.) WISTP 2010. LNCS, vol. 6033, pp. 17-30. Springer, Heidelberg
(2010)
29. Veyrat-Charvillon, N., Standaert, F.: Mutual Information Analysis: How, When
and Why? In: Clavier, C., Gaj, K. (eds.) CHES 2009. LNCS, vol. 5747,
pp. 429-443. Springer, Heidelberg (2009)
30. VLSI research group and TELECOM ParisTech: The DPA Contest (2008/2009),
http://www.dpacontest.org

A First-order Success Rate Results

We present here the results of the success rate metric on the two platforms
detailed in Sec. 5. First-order success rate is, for a given number of curves,
the probability that the correct key hypothesis is ranked first in the sorted
vector of hypothesis. These results are consistent with the guessed entropy metric
presented in Sec. 5.

1

CPA
CVM
GMIA
KDE
KNN
HE
BSE
CE
SPE
DCA

0.8

0.6

0.4

0.2

0

0 100 200 300 400 500 600

Number of curves

Fig. 3. First-order success rate on DPA Contest 2008/2009 curves of a DES