Social Recommendations Using Dynamic Profiles in Mobile Social Networks.

Mobile Social Networks (MSNs) enable their users to communicate in an ad-hoc infrastructure-less environment. Therefore, it is considered a very promising means of communication in cases of disasters and emergencies. However, because of the restricted radio coverage, network versatility and device power constraints, MSNs do not have continuous end-to-end connectivity between nodes. Information dissemination is one of the highly requested applications of MSNs. Multicast routing protocols are implemented to help disseminate data to target destinations. Their objective is to maximize the delivery ratio and reduce end-to-end latency. In this paper, directed multicast protocols are addressed, in which message forwarding is conditional on predefined criteria. One of the key factors to improve routing performance in MSNs is to carefully select candidate relays. In this paper, we exploit users' social profile and their interactions to improve relay recommendation and find target destinations. Our proposed protocol, Time-based Encounter of Socially Similar nodes, TESS, uses the social proximity between nodes taking into consideration the time at which nodes encounter. The proposed protocol reduces power consumption and network overhead, while increasing delivery ratios compared to related protocols. [ABSTRACT FROM AUTHOR]