Towards Trustworthy Peer-To-Peer Overlay Networks

Research Areas: Network Security,

Principal Investigator: Sanjay Rao; Cristina Nita-Rotaru

This project seeks to create trustworthy peer-to-peer overlay systems through fundamental advances to the state-of-the-art in the design of Internet-scale, overlay networks for performance-demanding applications in the presence of adversaries. Design limitations in popular peer-to-peer systems today may be exploited to cause large-scale denial of service attacks on nodes not even part of the overlay system. Further, adversaries may control the overlay construction to create a crippling impact on application performance. To tackle this, the project will (i) Design robust and attacker resilient adaptation protocols contributing to an emerging science of trustworthy adaptability that defines a new shift in building distributed systems; (ii) Obtain fundamental insights into the interplay between the design of group management algorithms and their vulnerability to being exploited to launch distributed denial of service attacks on the Internet; (iii) Explore the interactions between peer-to-peer design, resulting traffic characteristics, and implications for distinguishing normal peer-to-peer traffic patterns from anomalous ones; and (iv) Design mechanisms for reliable, scalable and adversary-resilient key dissemination to help ensure confidentiality and integrity of application-specific data.

The project will demonstrate and validate the novel proposed mechanisms in the context of mature and widely deployed peer-to-peer systems. Peer-to-peer video broadcasting will be used to promote online education in the Lafayette area, and to broadcast a security-related seminar series. The project will benefit the design of large-scale testbeds such as GENI. The PIs will communicate with developers of popular peer-to-peer systems to alert them to critical design vulnerabilities in their systems.

Keywords: trustworthy, peer-to-peer, overlay networks