MISG 2016 Graduate Workshop e-vota AIMS South Africa & SUN J. - - PowerPoint PPT Presentation

misg 2016 graduate workshop e vota
SMART_READER_LITE
LIVE PREVIEW

MISG 2016 Graduate Workshop e-vota AIMS South Africa & SUN J. - - PowerPoint PPT Presentation

1 MISG 2016 Graduate Workshop e-vota AIMS South Africa & SUN J. W. Sanders 2 The problem Election by vote: from clubs to nations. Many voting protocols are in common use and their features are well known (and perhaps


slide-1
SLIDE 1

1

✬ ✫ ✩ ✪

MISG 2016 Graduate Workshop e-vota

  • J. W. Sanders

AIMS South Africa & SUN

slide-2
SLIDE 2

2

The problem

Election by vote: from clubs to nations. Many voting protocols are in common use and their features are well known (and perhaps surprising at first). However all voting systems share important properties, including:

  • voter authentication
  • confidentiality of vote
  • accountability of result.

Can we design a distributed voting system?

slide-3
SLIDE 3

3

Characteristics of this problem

What characterises a typical MISG problem? How this problem is similar; and different. What is required: a design; its correctness and efficiency. What is not required: a program; testing it in various cases. The maths is pure, discrete and perhaps unfamiliar.

  • Abstraction. Design space. Nondeterminism.
slide-4
SLIDE 4

4

Specification

Assume the voting protocol is given by a ‘black box’ procedure. Concentrate on the rest of the system. Our system is specified by its functionality (reflects the voting protocol) extra features (authentication, confidentiality, anonymity, + security?, + robustness?)

slide-5
SLIDE 5

5

Design techniques

  • Describing a distributed design.

Modularity. Information flow by shared variables or message passing.

  • Reasoning about distributed behaviour.

Each module must be autonomous. Invariant properties.

  • Public key encryption.

Secure communication. Digital signatures.

  • Mathematical notation.

Z formalism.

slide-6
SLIDE 6

6

Example: the voting protocol

Voters Candidates Rankings := perms(Candidates) Votes := Voters → Rankings VProtocol := Votes → Rankings

slide-7
SLIDE 7

7

Concerns

  • 1. Modelling: how to abstract (what is ‘observable’?).
  • 2. Does the distributed e-format offer new possibilities for a voting

system?

  • 3. Correctness?
  • 4. Efficiency?
slide-8
SLIDE 8

8

References

  • Survey on electronic voting schemes, Laure Fouard, Mathilde

Duclos and Pascal Lafourcade. 65 pages.

  • Design and analysis of a practical e-voting protocol, Mari´

an Novotn´

  • y. 14 pages.
  • Analysis of an Electronic Voting System, Tadayoshi Kohno,

Adam Stubblefield, Aviel Rubin and Dan Wallach. 23 pages.