Datenbestand vom 13. Juni 2019

Warenkorb Datenschutzhinweis Dissertationsdruck Dissertationsverlag Institutsreihen     Preisrechner

WICHTIGER HINWEIS
DER VERLAG IST IN DER ZEIT VOM 12.06.2019 BIS 23.06.2019 AUSCHLIESSLICH PER EMAIL ERREICHBAR.

aktualisiert am 13. Juni 2019

ISBN 9783843933490

Euro 42,00 inkl. 7% MwSt


978-3-8439-3349-0, Reihe Informatik

Matthias Wichtlhuber
On Collaborative Mechanisms for Content Distribution

198 Seiten, Dissertation Technische Universität Darmstadt (2017), Softcover, A5

Zusammenfassung / Abstract

Content Delivery Networks (CDNs) deliver a large share of the global network traffic. Consequently, optimizing CDN technology is a significant lever to optimize the resource usage of the global Internet. However, doing so is challenging: CDN providers are embedded in a complicated environment of business relations and technical dependencies. More precisely, CDNs depend on Internet Service Providers (ISPs), content providers, and end-users. An optimization benefiting a CDN provider may hurt the interest of other stakeholders and may therefore lack the incentive to be applied. The need to align the economic perspective and the technical perspective makes the optimization of CDNs an interesting and challenging field at the intersection of computer science and economics.

Consequently, this thesis aims at the design and evaluation of collaborative mechanisms for content delivery. Collaborative mechanisms introduce interfaces between CDN providers and other actors of the content delivery environment. They are designed to meet two goals: (a) optimizing content delivery with respect to relevant performance indicators while (b) provisioning incentives to the stakeholders to participate in the collaborative mechanism. More precisely, this thesis contributes and evaluates two different collaborative mechanisms for CDN-ISP collaboration and CDN-end-user collaboration. Both mechanisms are shown to result in a win-win situation for the collaborating partners with respect to relevant cost and performance metrics.