Interest-aware implicit multicast (iCast)
Saeed Moghaddam, Ahmed Helmy · 2014
The unprecedented tight coupling between mobile devices and their users provides new opportunities to infer users' behavior and interest. Much of the future mobile services shall center on human behavior and interests, and will rely heavily on the paradigm of multicast communications, as in group communication in mobile social networks. In this paper, we introduce a new casting paradigm, interest-aware implicit multicast (iCast) that works based on the inferred interest profiles. In this paradigm, messages are sent to a behavioral interest profile (not to an IP or device address). In this manner it is considered opportunistic where the users are implicitly matched based on their profiles using co-clustering algorithm. There is no need for explicit member join or leave, and no per-group membership management. We explore a spectrum of architectures including centralized, semi- centralized and distributed and conduct a campus-wide case study for the evaluation of the second architecture. Within this architecture, we suggest two different message delivery approaches including semi and full interest-aware; and show how the later approach can significantly improve the performance utilizing a set of recommendations that are built based on a behavioral interest model of mobile society.