Rewarding Good Behavior in Peer-to-Peer Networks

Abstract

Large-scale human cooperation is vital to society. Designers and community facilitators of peer-to-peer networks are constantly investigating free-riding prevention measures for their platforms. Such schemes of rewarding good behavior remain a challenging problem in fully decentralized P2P networks. This work is a first attempt to integrate cutting-edge technologies developed by the Tribler team into a functional reputation-driven P2P network. We start by evaluating the possibilities of our system in supporting generous donations, and we continue with the issue of enforcing preferential treatment by implementing a reciprocity mechanism in Libswift, the reference implementation of the upcoming IETF Peer-to-Peer Streaming Internet Standard. Additionally, we use and evaluate solutions developed during this project in two experiments, followed by a third one whose goal is to demonstrate the concept of indirect reciprocity using the fully integrated system. Aiming to open the road for further research, we conclude with initial insights into the behavior of the involved components, alongside with a number of potential weaknesses.