Network infrastructures are being continuously challenged by increased demand, resource-hungry applications, and at times of crisis when people need to work from homes such as the current Covid-19 epidemic situation, where most of the countries applied partial or complete lockdown and most of the people worked from home. Opportunistic Mobile Social Networks (OMSN) prove to be a great candidate to support existing network infrastructures. However, OMSNs have copious challenges comprising frequent disconnections and long delays. we aim to enhance the performance of OMSNs including delivery ratio and delay. We build upon an interest-aware social forwarding algorithm, namely Interest Aware PeopleRank (IPeR). We explored three pillars for our contribution, which encompass (1) inspect more than one hop (multiple hops) based on IPeR (MIPeR), (2) by embracing directional forwarding (Directional-IPeR), and (3) by utilizing a combination of Directional forwarding and multi-hop forwarding (DMIPeR). For Directional-IPeR, different values of the tolerance factor of IPeR, such as 25% and 75%, are explored to inspect variations of Directional-IPeR. Different interest distributions and users’ densities are simulated using the Social-Aware Opportunistic Forwarding Simulator (SAROS). The results show that (1) adding multiple hops to IPeR enhanced the delivery ratio, number of reached interested forwarders, and delay slightly. However, it increased the cost and decreased F-measure hugely. Consequently, there is no significant gain in these algorithms. (2) Directional-IPeR-75 performed generally better than IPeR in delivery ratio, and the number of reached interested forwarders. Besides, when some of the uninterested forwarders did not participate in messages delivery, which is a realistic behavior, the performance is enhanced and performed better generally in all metrics compared to IPeR. (3) Adding multiple hops to directional guided IPeR did not gain any enhancement. (4) Directional-IPeR-75 performs better in high densities in all metrics except delay. Even though, it enhances delay in sparse environments. Consequently, it can be utilized in disastrous areas, in which few people are with low connectivity and spread over a big area. In addition, it can be used in rural areas as well where there is no existing networks.
Computer Science & Engineering Department
MS in Computer Science
Prof. Sherif Aly
Prof. Soumaia Alayyat
Committee Member 1
Prof. Tamer Elbatt
Committee Member 2
Prof. Amr El Mougy
Institutional Review Board (IRB) Approval
Approval has been obtained for this item
Shahin, Y. S.
(2021).Enhanced Interest Aware PeopleRank for Opportunistic Mobile Social Networks [Master's Thesis, the American University in Cairo]. AUC Knowledge Fountain.
Shahin, Yosra Saad. Enhanced Interest Aware PeopleRank for Opportunistic Mobile Social Networks. 2021. American University in Cairo, Master's Thesis. AUC Knowledge Fountain.