..................................................................................................................................................................... 4 BACKGROUND ................................................................................................................................................................. 4 METHODS ........................................................................................................................................................................ 4 RESULTS ......................................................................................................................................................................... 4 INTRODUCTION ............................................................................................................................................................ 5 BACKGROUND AND MOTIVATION ................................................................................................................................... 5 PROBLEM STATEMENT .............................................................................................................................................. 6 OVERVIEW ...................................................................................................................................................................... 6 RESEARCH QUESTION/HYPOTHESIS ................................................................................................................................ 6 OBJECTIVES AND AIMS .............................................................................................................................................. 7 OVERALL OBJECTIVE ...................................................................................................................................................... 7 SPECIFIC AIMS ................................................................................................................................................................ 7 BACKGROUND AND SIGNIFICANCE ....................................................................................................................... 8 RESEARCH DESIGN AND METHODS ....................................................................................................................... 9 PROTOTYPE ..................................................................................................................................................................... 9 ENCRYPTION/ DECRYPTION OF BALLOTS ....................................................................................................................... 12 Elgamal .................................................................................................................................................................... 12 Initialisation and Key generation: ....................................................................................................................................... 12 Encryption .............................................................................................................................................................................. 12 Decryption ............................................................................................................................................................................. 13 Distributed private-key ......................................................................................................................................................... 13 Key-generation and initialisation with distributed private-key (Damgård, 2002): ........................................................ 13 Choose of cyclic group an key length ................................................................................................................................ 13 BALLOT COUNTING – HOMOMORPHIC ENCRYPTION ....................................................................................................... 15 Encryption ................................................................................................................................................................ 16 Summation ............................................................................................................................................................... 16 Decryption ............................................................................................................................................................... 16 Shifted summation .................................................................................................................................................... 16 CONCLUSION ............................................................................................................................................................... 18 REFERENCES ............................................................................................................................................................... 19 LIST OF FIGURES ........................................................................................................................................................... 19 ABSTRACT Background In the past few years, there were a lot of pilot projects to test whether and how secure eVoting can be. For example the cantons Zürich, Neuchatel and Geneva took the lead and started an eVoting system for communal and cantonal elections and ballots. Today only Geneva and Neuchatel still has their eVoting system since the other cantons faced too much open (security) questions. After having a closer look how the eVoting system of Geneva works, we still think that the enand decryption of the ballots can be done in a better and more secure way. So this prototype suggests another method for ballot en/ decryption which will be more secure.Background In the past few years, there were a lot of pilot projects to test whether and how secure eVoting can be. For example the cantons Zürich, Neuchatel and Geneva took the lead and started an eVoting system for communal and cantonal elections and ballots. Today only Geneva and Neuchatel still has their eVoting system since the other cantons faced too much open (security) questions. After having a closer look how the eVoting system of Geneva works, we still think that the enand decryption of the ballots can be done in a better and more secure way. So this prototype suggests another method for ballot en/ decryption which will be more secure. Methods We will create a prototype with an optimized ballot en/ decryption and an anonymous counting system which increases the privacy and security of the whole eVoting process. Results Prototype in PHP and Java which will clearly demonstrate how a secure en/ decryption and anonymous counting system can be implemented. INTRODUCTION Background and Motivation Some years ago, several cantons in Switzerland started to introduce an eVoting system. Most of them were pilot projects but in Geneva and Neuchatel you can still vote online. Also the canton Berne started to provides an eVoting system for cantonal elections and ballots. But they did not implement an new solution but used the one from Geneva. This means that the ballot from Berne is sent to Geneva where the ballot will be analysed and evaluated. Since the Swiss government would like to provide an eVoting system to all citizens, everyone is affected by potential security leaks of the system. As we have seen, the system form Geneva is very popular and successful, so we would like to have a closer look to one potential security leak and provide a solution. PROBLEM STATEMENT Overview First of all, the Geneva eVoting (Geneva) system is not fully documented. This is from our point of view, a first indication that the system is not perfectly secure what also reduces the confidence level. There is a short description (in French) of the procedure which already makes the reader feel unconfident about the privacy level. The procedure is as follows: 1. The ballot is encrypted 2. On the server, the ballot is decrypted and checked whether it is valid or not 3. If the ballot is valid, it is again encrypted and placed in a electronic ballot-box 4. The analysis of the ballots needs a master-password which decrypts all ballots and counts them. 5. The result of the ballot can be published. As we can see, there is already a potential security leak in this short description. If some knows the master-password, then this person can control the result of the ballot. Of course there is a lot of security to protect this master-password from abuse, but still there is one person that can access it. If everyone trust this person everything is fine, but how to react if someone complains about the fairness and thinks that this person changed the result of the ballot? Another problem is that each ballot is decrypted when it reaches the server to control it. There you could filter the ballots and also change the result. Furthermore to calculate the result, again all ballots must be decrypted, and if someone manages to create a link between the ballot and the elector the privacy is gone. Research Question/Hypothesis There must be a solution where no master-password is needed to count the ballots and publish the result of the ballot in a way that ensures complete privacy and fairness. Furthermore the ballot should not be decrypted during the counting process. This could be done by using an asymmetric encryption process called elgamal and a distributed private-key system to split up the master-password. We will create a prototype which shows that this encryption works in real-life applications even though it has already been proven in academic context. OBJECTIVES AND AIMS Overall Objective The main objective of this project, was to show that eVoting can be made secure and transparent. In order to make the user of the system sure that what he