The secure platform problem taxonomy and analysis of existing proposals to address this problem

Abstract

One of the main open issues in electronic government is the fact that the individual users' multi-purpose computing platforms are used. In terms of security, no guarantee is given since these platforms are not under the government authority's control. Even worse, the number of malware infected computing platforms increases. This so-called Secure Platform Problem and approaches aiming to solve it are objects of investigation in this work. We define criteria that need to be ensured to address this problem. Furthermore, we propose a taxonomy to classify existing approaches. Based on the classification and the criteria, we analyze the different types of approaches by providing concrete examples. Hereby, we show that none of the existing approaches fully meets our criteria. Thereby, we focus on the most security critical class of electronic government services, namely electronic voting over the Internet. However, most of the discussed approaches as well as the criteria and classification can also be applied to other governmental applications.

A Complete Characterization of Secure Human-Server Communication

“…Recall that Rule (1) is the only rule producing fresh constants and thereby creating Fr facts. By Equation (23), every fresh constant is produced at most once in a trace. Fresh constants can be obtained (generated) by honest agents using Rule (2).…”

“…Most existing work focuses on particular scenarios, for instance on browserbased security protocols [12], [13], login procedures [14], solutions for online banking [27], or Internet voting [23]. A general approach to modeling and reasoning about such systems are security ceremonies [9].…”

Voter, What Message Will Motivate You To Verify Your Vote?

“…A message utilizing images, sound or video would be very effective but would have to be delivered via a supporting platform. Unfortunately there is always the possibility that the voting platform has been compromised [50], thus delivering instructional messages this way seems insecure 2 . Hence we designed messages which could easily be delivered via a secure channel (e.g.…”