A Decentralized and Distributed E-voting Scheme Based on - - PowerPoint PPT Presentation

a decentralized and distributed e voting scheme based on
SMART_READER_LITE
LIVE PREVIEW

A Decentralized and Distributed E-voting Scheme Based on - - PowerPoint PPT Presentation

A Decentralized and Distributed E-voting Scheme Based on Cryptographic Shuffles Decentralized and Distributed Systems Laboratory Andy Caforio Responsible: Prof. Bryan Ford Supervision: Linus Gasser, Philipp Jovanovic 1 Way back when


slide-1
SLIDE 1

A Decentralized and Distributed E-voting Scheme Based on Cryptographic Shuffles

Decentralized and Distributed Systems Laboratory Andy Caforio Responsible: Prof. Bryan Ford Supervision: Linus Gasser, Philipp Jovanovic

1

slide-2
SLIDE 2

Way back when…

2 [https://heliosvoting.org/]

slide-3
SLIDE 3

Helios

  • Started in 2008
  • First web-based, verifiable e-voting scheme
  • Leverages cryptographic shuffles

3

slide-4
SLIDE 4

Helios - Features

  • Auditable elections

○ Encryption proof ○ Shuffle proof (Sako-Kilian) ○ Decryption proof

  • User authentication
  • Front- and back-end implementation

4

slide-5
SLIDE 5

Helios - Verifiability

  • Users can verify that their vote was counted
  • Shuffle weeds out malicious servers
  • Honest servers will perform decryption

5

slide-6
SLIDE 6

Helios - Protocol

1. Cast 2. Publish 3. Shuffle 4. Audit 5. Decrypt 6. Tally

6

slide-7
SLIDE 7

Helios - Disclaimer

  • Helios does not enforce anonymity
  • Voters may be subject to coercion

7

slide-8
SLIDE 8

Helios - Downsides

  • Centralized
  • Conventional database storage
  • Very slow shuffles

8

slide-9
SLIDE 9

Helios - Improvements

  • Centralized Cothority
  • Conventional database storage Skipchains
  • Very slow shuffles Neff

9

slide-10
SLIDE 10

Back to the future

10

slide-11
SLIDE 11

Protocols - DKG

  • Distributed Key Generation
  • Create public/private key pair
  • Split private key
  • Part of kyber library

11

slide-12
SLIDE 12

Protocols - Neff Shuffle

  • Novel verifiable shuffle concept by Andrew Neff
  • Orders of magnitude faster than Sako-Kilian scheme

12

slide-13
SLIDE 13

Protocols - Decryption

  • After election termination and audit
  • Reconstruct plaintext ballots with shared secret keys
  • Cannot be done by a single node

13

slide-14
SLIDE 14

14

Storage

  • Master

○ System configurations ○ List of admins, roster etc.

  • Link

○ Reference to election skipchain

  • Election

○ Settings ○ DKG public key, list of voters etc.

  • Ballot

○ Casted vote (one per block)

  • Shuffle

○ Permuted and re-encrypted ballots

  • Decrypt

○ Ballot plaintexts

slide-15
SLIDE 15

Practical

  • Go implementation
  • Built on top of cothority and kyber
  • Protobuf API

15

slide-16
SLIDE 16

Benchmarks - Helios

  • Shuffle of 500 ballots ~130s
  • 2.2 GHz dual core machine

16 [Ben Adida. Helios: Web-based open-audit voting]

slide-17
SLIDE 17

Benchmarks

17

  • Shuffle
  • 1.4 GHz dual core
  • Real world context?
slide-18
SLIDE 18

Overview

  • Distributed e-voting scheme
  • Improves on Helios

○ Distributed ○ Faster

  • Built on top of DEDIS infrastructure

18

slide-19
SLIDE 19

Gory details

  • Cryptographic background

○ Framework (elliptic curve etc.) ○ Shuffles ○ Verifiability

  • Protocols

○ Networking

  • Usage

○ Authentication ○ Front-end

19

slide-20
SLIDE 20

References

  • Repository: https://github.com/dedis/student_17_evoting
  • Report: https://github.com/dedis/student_17_evoting/blob/master/report.pdf

20