Content Distribution

When users encounter with each other through opportunistic contacts, they may exchange content by using mobile devices directly, without network infrastructures. To distribute content from the source user to the destination, the appropriate relay nodes should be selected to deliver content, with considerations such as bandwidth utilization, interval time, interest in the content, etc. [40-46]. Besides, to further improve the QoE, the environment aware is also needed in MSNs to analyze and predict users’ social features, based on users’ environment information (e.g. location, conversation time etc.) [47-51].

To efficiently distribute content, Wu et al. [51] study the social features of users during the information propagation in MSNs. Two classes of selfishness are considered which are individual selfishness and social selfishness, respectively. Xu et al. [52] propose an incentive scheme to select the relay node in MSNs to overcome the problems caused by the users’ selfishness. The problem of selection is formulated as a bargain game and the Nash equilibrium is obtained to incentive users to relay the content. Mei et al. [44] propose two forwarding protocols for MSNs which are Give2Get epidemic forwarding and Give2Get delegation forwarding, respectively. In the two protocols, the number of copies of data is limited to reduce the total overhead.

Bulut et al. [53] consider the friendship between users in MSNs and define a new metric to find the direct and indirect friends. Based on the friendship information in MSNs, the authors propose a friendship-based routing scheme for the content delivery. Xu et al. [54] study the epidemic information dissemination in MSNs and develop an analytical model to analyze dissemination process. In the model, the authors adopt preimmunity and immunity to represent the attitude about whether the user is interested in the information or not. Wang et al. [55] propose a cloud-based multicast scheme for MSNs, in which the information forwarding process is divided into two phases which are pre-cloud and inside-cloud.

< Prev   CONTENTS   Source   Next >