Data Forwarding Using Interest Relationship in Delay Tolerant Social Networks

Abstract

e devices and various types of wireless networks. This trend brings an opportunity for individual to access digital contents, share resources and communicate any where any time facilities. However, the connectivity coverage rate did not go hand in hand to the trends of mobile phone penetration rate, and there is connectivity coverage disparity among developed countries and developing ones. This motivates that infrastructureless networks such as Delay Tolerant Networks(DTNs)to play an alternative communication approach and to be leveraged in different application scenarios. The characteristics of this kind of networks (i.e., intermittent connectivity and transient end to end path) create a challenging problem in routing data. For example, to deliver a given message towards destination node/groups of nodes, the forwarding scheme uses a store-carryforward and utilizes different heuristics to select a potential forwarder node/intermediate node. This thesis focuses on those routing schemes that consider social properties of individual nodes, and aims to design forwarding algorithms as a means to improve network performance. Existing works overlooked interest relationships among users to enhance forwarding performance. Therefore, the main objective of the study is: designing social-aware based opportunistic forwarding solutions to attain a better forwarding performance. Therefore, Interest Relationship Based Forwarding Scheme in Delay Tolerant Social Networks (DTSNs) named as IntReF is proposed. It is motivated by the fact that combining two or more social similarity in terms of user interest, and its relationships in accordance with human interest dynamics showed fat-tailed probability distributions in human daily activities. To evaluate the performance, an extensive simulations experiment is conducted. IntReF is compared against two selected schemes Bubblerap and Epidemic. The performance results shows IntReF achieves a better forwarding performance in terms of delivery ratio, average latency, as result compared to bubblerap and Epidemic. The simulation results demonstrate that IntReF outperforms bubblerap and Epidemic with higher message delivery ratio and maintains a reduced average latency. Moreover, IntReF can outperforms around 24.66% delivery ratio in average in terms of varied value of buffer size, message time to live (TTL) and simulation duration parameters. Simulation evaluation results towards varied value of buffer size, message time to live (TTL) and simulation duration support the effectiveness of the scheme