Clique-aware mobile social clouds

The important role played by cliques in identifying cohesive subgroups of people has been theorized and explored by sociologists years ago, but only recently investigated in large-scale social networks. In this paper we focus on the interplay between cliques established by on-phone communications and the urban locations their members share each other. The results about co-located cliques has been achieved through the extensive analysis of a large anonymized dataset of Call Detail Records (CDR) relying on the phone activities of nearly 1 million people in the city of Milan. Taking the cue from the observation of cliques, the paper envisions and designs a novel clique-support service for mobile users by fully exploiting the current virtualization process that is radically transforming the core network of mobile operators. The approach we propose brings together a few important contributions: first, it concretely shows that the current NFV-enabled trend of placing cloud services at the edge of the operator's network is viable and may have a payoff in terms of traffic offloading and improved user's experience; secondly, it demonstrates for the first time that a few typical cloud-based services can effectively be directly performed inside the mobile network by simply leveraging the rich amount of data about users' location and mobility behavior.