Towards a privacy-preserving hybrid radio network: design and open challenges

A large-scale system hundreds of millions of people encounter every day is radio. While this “system” has a tremendous reach it also is technologically outdated. Technological constraints offer great protection of listeners' privacy but prevent radio stations from implementing modern business models like personalization and mining valuable user information, at the same time. This paper describes a fully distributed system that aims at overcoming current technological constraints in interconnecting radio stations and their listeners while retaining a comparable protection of sensitive user data. Namely, a peer-to-peer architecture with integrated data-mining-capabilities employing differential privacy is proposed. The system will offer personalizable radio programs to listeners and it will enable radio stations to gather valuable information about their listeners. Furthermore, this paper points out key challenges in deploying, bootstrapping and maintaining such a distributed system.