Device-to-Device Mobile Data Offloading for Music Streaming

Device-to-device communication (also referred to as opportunistic networking) is considered a feasible means for offloading mobile data traffic. Due to the sporadic nature of contact opportunities, applications in the domain of device-todevice communication are assumed to be delay-tolerant, with content delivery deadlines being in the order of hours. However, predictions suggest that by 2020 more than 75% of the traffic volumes at mobile operators will be generated by multimedia contents which is often seen as data served in real-time. In this paper we explore how the concept of opportunistic networking can be used for dissemination of real-time streaming contents for users in urban environments without degrading quality of experience. We first present a general framework for offloading multimedia data that is organized in terms of playlists, and we then investigate the performance of the framework in realistic urban environments using the music streaming service Spotify as a use-case. Our results show that it is feasible to use opportunistic device-to-device communication in the context of music streaming. We demonstrate that the system performance is insensitive to a number of parameters such as playlist length distribution, and initial content availability distribution, however it exhibits sensitivity towards the amount of requested data and the node density.