Greetings from the Australian national University where Vanessa Teague ,University of Melbourne is talking on "End-to-end verifiable cryptographic voting schemes". She reviewed the various electronic voting systems which have been used in Australia, such as "iVote" used in the NSW 2011 state election. The iVote system had obvious flaws, as there is limited evidence trail. A new version is proposed for the 2015 NSW election. The process is described in "iVote Strategy for the NSW State GeneralElection 2015" (NSW Electoral Commission, August 2013). What is remarkable about this is not the process proposed is flawed, but how the NSW Electoral Commission could propose such an obviously flawed system on a second attempt.
I have used the EVACS system, developed by the local company Software Improvements, to vote in Australia is the Australian Capital Territory (ACT) in the last two local elections. Also I have used a pilot system for one federal election. Dr Teague described EVACS as the only open source system, well thought out and extensive analysis. The audience pointed out that all the software may not be open source (Evoting pioneer plays politics with open source, Steven Deare, LinuxWorld, 11 August, 2004).
Dr Teague pointed out that as good as it is, even eVACS does not provide a way to verify the vote counting is correct. Cryptographic techniques can be used to do this, but it may not be practical. The system has the voter receiving an encrypted receipt of their vote. The voter can check their recipient against a list published by the voting system.
The encrypted system relies on the list of candidates being randomly listed for each voter. The list of what the voter voted for can then be published, without revealing who was voted for. Unfortunately there was not sufficient time in the talk to explain the detail of how this works.
One of the problems with any e-voting polling place system is the cost of the hardware. For the Australian local, state and federal systems a simple change would have it cheaper. The ACT and federal elections have e-voting at the larger polling palces used for pre-polling. Normally the pre-polling is just used for people who are unable to vote on polling day, but if citizens were encouraged to use this system, them most of the votes could be collected with a few electronic systems (and the other votes collected on paper).
There is a paper available "Verifiable Postal Voting", Josh Benaloh, Peter Y. A. Ryan, Vanessa Teague in Security Protocols XXI
No comments:
Post a Comment