We utilize the inherent fraud-resilience of proximity to stop unlimited fake identity creation

Close proximity is a property which is easy to validate. Also online it is possible to prove that somebody is less than 500km away. This enables us to distinguish close acquaintance from far away strangers.

Simple methods such as sending a large number and expecting it to be returned quickly offers a tamper-proof proximity measure. The laws of physics and limited speed of light prevent tampering and can prove that you are located nearby.

This master thesis will create a mathematical model which reflects the success of proximity-based defenses for a given strength of the Sybil attack.

We make a realistic assumption on attacker capabilities. It is far harder to create fake identities in close proximity to a victim than at any other random location on the Internet. We assume that attackers have finite resources and some financial constraints. The goal is to prove strong mathematical properties of defenses and attacks. In a generic cooperative agent model build trust after several successful interactions. We believe that a successful strategy to build cooperative cluster in an overwhelmingly defectors-dominated world is to do some initial risk taking when meeting strangers. By adding our proximity mechanism it is possible to build highly effective mechanisms to identify cooperators in close proximity plus test, discard non-cooperators, and mitigate (Sybil) attackers. We further assume cooperators can gossip about the collaboration levels of others using a tamper-proof data structure, for instance, our IETF Internet Standard draft trustchain.

Seuken proved that it is impossible to defend against the Sybil attack in the general case, in any system which has single-report responsiveness. This thesis work is meticulously designed to bypass that impossibility result and create a meaningful, effective and even highly efficient system without single-report responsiveness. We have looked into restricting communication to strangers using proximity for years, #2541. Prateek Mittal from Princeton exploited the temporal dynamics of the Sybil attack for a defense. This work also builds upon our our prior work by Pim Otte.

Desired outcome:

• literature scan (games on graphs, All Novak work, social capital accounting, etc.)
• Sybil-Proofness using proximity in model form.
• ToDo: beautiful math proof

(still tentative, we just see how far we get with this master thesis direction)