A New Scalable Quality Guaranteed Model For The Real-time Multimedia Communications Across The InternetSherine Mohamed Abd El-Kader
AbstractIn this thesis, a new proposed model for the real-time multimedia communications across the Internet has been implemented to guarantee the voice/video required quality based on the available network resources (bandwidth and buffers). Basically, the new proposed model depends on simple Differentiated Services blocks at the edge of the networks. Although, it is simple from the design point of view, it is somehow complex from the inputs analysis point of view. This model claims that to guarantee the voice/video quality of service through the Internet, the used voice/video encoders should be analyzed carefully. But, the QBone (Quality of service Backbone) model does not analyze the input traffic (i.e. more simpler than the new proposed model) and reserves the path virtually between the source and the destination by the input traffic peak rate. The new proposed model has some calculated approximations for both the required buffers and bandwidth (it should be less than the peak and depends on the resources availability) to have the required acceptable quality. In fact, the QBone model does not success (due to its high costs) from the practical implementation point of view and most of the users are not wtlling to pay much money to have voice/video sessions across the Internet. So, the new proposed model depends not only on having network resources savings (may reach up to 74% more than the QBone model) but also having an acceptable guaranteed quality for voice/video sessions over IP based networks. In this thesis, 0.723.1 (5.3;6.4), G.711, G.726 (voice encoders), SQCIF (FIB), and QCIF (FIB) (video picture formats) characteristics representing the proposed model inputs have been analyzed carefully by implementing some capturing experiments. Then three scenarios (high, average, and low load scenarios) assumptions, descriptions, and their analytical representation have been described in details. In addition, the proposed model results, for two different configurations, have been obtained based on the packet delay, routers buffers consumptions, maximum jitter, and probability of packet loss parameters. Finally, a comparison between the new proposed model and the QBone model has been presented in details.
|Other Titles||نموذج جديد قابل للانتشار وضامن لجودة اتصالات الوسائط المتعددة ذات الزمن الحقيقي عبر شبكة الإنترنت||Issue Date||2003||URI||http://research.asu.edu.eg/handle/12345678/47805|
Recommend this item
Items in Ain Shams Scholar are protected by copyright, with all rights reserved, unless otherwise indicated.