On leveraging social relationships for decentralized privacy-preserving group communication

Millions of people use the Internet today for data exchange of sensitive information. A typical exchange usually occurs within the context of a group of users and often takes place in a centrally managed infrastructure such as Facebook. Evidence has shown that such systems are unable to properly protect the privacy of its users. Following this observation, we present a framework to support privacy-preserving group communication. Our framework is completely decentralized and leverages the social relationships between users to bootstrap a communication overlay whose robustness is improved with the addition of extra privacy-preserving links. The framework supports multiple groups in a scalable manner and defines mechanisms to handle churn in group membership. Finally, we also outline high-level protocols for a privacy-preserving micro-news application.