Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Date of Award

Spring 2013

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Department of Computer Science

Abstract

The act of casting a ballot during an election cycle has been plagued by a number of problems, both intrinsic and extraneous. The old-fashioned paper ballot solves a number of problems, but creates its own. The clear 21st Century solution is the use of an automated electronic system for collection and tallying of votes, but the attitude of the general populace towards these systems has been overwhelmingly negative, supported in some cases by fraud and abuse. The purpose of this thesis is to do a broad survey of systems available on the market now (both in industry and academia) and then compare and contrast these systems to an “ideal” system, which we attempt to define. To do this we survey academic and commercial literature from many sources and selected the most popular, current, or interesting of the designs—then compare the relative strengths and weaknesses of these designs. What we discovered is that devices presented by industry are not only closed-box (which makes them inherently untrustworthy), but also largely inept in security and/or redundancy. Conversely, systems presented by academia are relatively strong in security and redundancy, but lack in ease-of-use or miss helpful features found on industry devices. To combat these perceived weaknesses, we present a prototype of one system which has not previously been implemented, described in Wang [1]. This system brings together many ideas from academia to solve a significant number of the issues plaguing electronic voting machines. We present this solution in its entirety as open-source software for review by the cryptographic and computer science community. In addition to an electronic voting implementation this solution includes a graphical user interface, a re-encryption mix network, and several decryption methods including threshold decryption. All of these items are described in-depth by this thesis. However, as we discuss in the conclusion, this solution falls short in some areas as well. We earmark these problem areas for future research and discuss alternate paths forward.

Share

COinS