We adapt game theoretic methods for studying the security of two evoting systems: the Estonian E-Voting System (EstEVS) and Secure Electronic Registration and Voting Experiment (SERVE) performed in the United States of America. While these two systems are quite similar from technical side, security experts have made totally different decisions about their security--EstEVS was indeed used in practical elections while SERVE was decided to be insecure. The aim of this work is to clarify if the minor technical differences between these two systems were indeed a sufficient reason to distinguish between their security. Our analysis is oriented to practical security against large-scale attacks. We define a model for the real-life environment in which voting takes place and analyze the behavior of adversaries. We show that in our model EstEVS is secure and SERVE is not. The reliability of the results is still questionable because of our limited knowledge about many of the parameters. It turns out though that our main results are quite robust with respect to the choice of parameters.
[1]
D. Jefferson,et al.
Security analysis of SERVE 1 A Security Analysis of the Secure Electronic Registration and Voting Experiment ( SERVE )
,
2004
.
[2]
Dan S. Wallach,et al.
Analysis of an electronic voting system
,
2004,
IEEE Symposium on Security and Privacy, 2004. Proceedings. 2004.
[3]
Bruce Schneier,et al.
MODELING SECURITY THREATS
,
1999
.
[4]
Daniel E. Geer,et al.
Information Security: Why the Future Belongs to the Quants
,
2003,
IEEE Secur. Priv..
[5]
Jan Willemson,et al.
Rational Choice of Security Measures Via Multi-parameter Attack Trees
,
2006,
CRITIS.