Social Forwarding in Mobile Opportunistic Networks: A Case of PeopleRank

Abstract

The proliferation of powerful portable devices has created a new environment for networking. In such environment, devices are able to communicate between “challenged” networks such as sensor networks, mobile ad-hoc networks, or opportunistic ad-hoc networks using a set of protocols designed to accommodate disconnection. In particular, forwarding in mobile opportunistic networks needs to deal with such disconnections, and limited resources. As opposed to conventional communication that relies on infrastructure, these devices can use hop-by-hop opportunistic data forwarding between each other. In this environment, a device should decide whether or not to transfer a message at the time it meets another one. How to optimally select the next hop towards the destination in a way to minimize delay and maximize success rate is so far unknown. In opportunistic networks, a device has to decide whether or not to forward data to an intermediate node that it encounters. In this chapter, we describe PeopleRank as systematic approach to the use of social interaction as a means to guide forwarding decisions in an opportunistic network. PeopleRank ranks nodes using a tunable weighted combination of social and contact information. It gives higher weight to the social information in cases where there is correlation between that information and the contact trace information. More specifically, PeopleRank is an opportunistic forwarding algorithm that ranks the “importance” of a node using a combination of social and contact-graph information.