Exploiting Time-Varying Graphs for Data Forwarding in Mobile Social Delay-Tolerant Networks

Dongliang Xie,Xin Wang,Lanchao Liu,Linhui Mae
DOI: https://doi.org/10.1109/iwqos.2016.7590436
2016-01-01
Abstract:With the rapid shift from end-to-end communications to content-based data sharing, there are increasing interests in exploiting mobile social Delay-Tolerant Networks (social DTNs) to deliver data, where the forwarding decision is usually made by comparing the social metrics of encountered nodes. Existing studies mostly derive long-term statistical social metrics without considering the temporal impact from node mobility. We exploit the time-varying contact graphs to analyze the dynamics of social DTNs based on two groups of datasets. Based on the analysis, we derive the time-varying characteristics of node contacts, durative and periodicity, and apply them to more accurately predict the corresponding time-varying social metrics (TSMs). We further propose a two-stage opportunistic forwarding strategy to select relays based on TSMs. Our simulation results verify the importance of the two properties we observe and the effectiveness of our algorithm in tracking time-varying social metrics. We also show the potential of our algorithm in finding general time varying metrics to improve the data dissemination performance of other opportunistic forwarding schemes.
What problem does this paper attempt to address?