Tezin Türü: Yüksek Lisans
Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Enformatik Enstitüsü, Bilişim Sistemleri Anabilim Dalı, Türkiye
Tezin Onay Tarihi: 2018
Tezin Dili: İngilizce
Öğrenci: SERHAT AĞIRBAŞ
Asıl Danışman (Eş Danışmanlı Tezler İçin): Altan Koçyiğit
Eş Danışman: Cüneyt Sevgi
Özet:Constant bit rate, time sensitive data delivery is needed for many network applications and these applications usually require high throughput and less variable delay. When such applications run on mobile devices, the bandwidth available and the other characteristics of the primary network connection may not be sufficient to provide necessary quality of service. On the other hand, most of the mobile devices are multihomed that is they are equipped with more than one network interface, hence they can be connected to more than one network simultaneously. Therefore, the bandwidth aggregation is a viable option for better quality of service in such cases. This thesis tackles constant bit rate data delivery problem by utilizing multiple network connections to satisfy bandwidth requirements in a cost effective manner specifically for time sensitive file streaming applications. The proposed method, called Cost Aware TCP Scheduler (CATS), aggregates the resources of two network connections which are available on the client device in a cost aware manner to deliver enough bandwidth for streaming applications. One of the connections is called the free connection and it is considered to have fluctuating throughput and variable delay but no monetary cost of use. The second connection is the paid connection that provides higher throughput and less variable delay but there is monetary cost associated with data transfer. CATS schedules data transfer over these connections to ensure timely delivery of data while minimizing the data transfer cost. Experimental results show that CATS maximizes the utilization of the free connection and minimizes the utilization of the paid connection thereby reducing the total monetary cost without causing significant quality degradation.