English

Vid tillräcklig efterfrågan

Content Delivery Networks (CDNs) have emerged as a new paradigm to improve users' quality of experience (QoE) and to reduce the cost of content delivery by pushing contents from the network core to the customer edge. This course is aimed at providing the doctoral students with fundamental knowledge about the CDN paradigm and its main tasks, architectures and the related open challenges that can be seen as interesting research topics. The course introduces formal methods to be exploited for the investigation of CDN performance issues with respect to a variety of performance objectives such as energy consumption, bandwidth usage, and content delivery time.

1. Introduction: taxonomy of the internet contents and trends, the new challenges of user-generated contents, content delivery tasks (content management and request routing), traditional solutions to content delivery (web-based content delivery methods).
2. Overview of CDN architectures: overlay CDNs, Peer-to-Peer (P2P) CDNs, hybrid architecture, Content Centric Networks (CCNs).
3. Overlay CDNs: content management, request routing, load balancing, challenges and open questions.
4. P2P CDNs: structured and unstructured P2P architectures, search methods, replication and caching, delivery methods, applications, challenges (mobility, ISP traffic imbalance).
5. CCN: in-network caching techniques, practical challenges (feasibility with respect to the conventional Internet architecture).
6. Special topics in video delivery networks: video streaming protocols, video delivery architectures, current issues and open questions.
7. Performance modelling and analysis: performance metrics (response time, bandwidth usage, and energy consumption), mathematical modelling and analysis, guidelines for efficient design (capacity planning, dynamic CDNs and cooperative content management).
8. Future trends: CDN as a service (cloud CDNs).

demonstrate knowledge about the fundamental characteristics of various content types such as delay requirement and the amount of resources, such as bandwidth, they consume for distribution.
demonstrate a thorough understanding of CDN tasks and the prevalent overlay architectures and technologies.
show knowledge about the strengths and weaknesses of different CDN architectures and the situations where a given overlay architecture is preferred over another one.
be able to indicate knowledge about the formal methods, especially from optimization theory, that can be employed to formulate and express content distribution problems.

demonstrate the theoretical skills required for the mathematical modelling of CDNs with regards to different, sometimes conflicting, objectives.
demonstrate the basic skills required for the design of (centralized and distributed) methods/algorithms facilitating efficient content management and delivery.
conduct a thorough literature survey on the assigned topic(s) and share their findings with the peer participants.

be able to evaluate the QoE requirements and delivery costs of different content types.
be able to assess the architectural and functional drawbacks of the existing CDN architectures and the possible ways to enhance these architectures.
be able to compare the existing solutions for content management and delivery, and evaluate the efficiency of each solution for particular content types.
be able to assess the performance of the existing content delivery solutions in face of new trends such as user-generated contents.

Föreläsningar

Projekt

Skriftlig rapport

Inlämningsuppgifter

Seminarieföredrag av deltagarna

The doctoral students are also required to do a mini-project as part of course requirements.

Underkänd, godkänd

i) be familar with the Internet architecture and protocols, ii) have basic mathematical skills of modeling optimization problems, and iii) be familiar with distributed computing techniques.

Pathan, M., Sitaraman, Ramesh K. & Robinson, D.: Advanced Content Delivery, Streaming, and Cloud Services. Wiley, 2014.
Buyya, R., Pathan, M. & Vakali, A.: Content delivery networks. Springer, 2014.

In addition, scientific papers

Course coordinator: Saeed Bastani, saeed.bastani@eit.lth.se

EIT130F

 -08-25

FN1/Anders Gustafsson