Today, a large fraction of Internet traffic is originated by Content Delivery Networks (CDNs). To cope with increasing demand for content, CDNs have deployed massively distributed infrastructures. These deployments pose challenges for CDNs as they have to dynamically map end-users to appropriate servers without being fully aware of the network conditions within an Internet Service Provider (ISP) or the end-user location. On the other hand, ISPs struggle to cope with rapid traffic shifts caused by the dynamic server selection policies of the CDNs.

The challenges that CDNs and ISPs face separately can be turned into an opportunity for collaboration. For CDNs, a closer collaboration with an ISP is an important step towards improving their competitiveness in the content delivery market. For ISPs, it is a step towards better management of the traffic that flows within their networks. ISP-CDN collaboration is also an enabler for both ISP and CDN to jointly introduce new applications to end-users by utilizing the global footprint of a CDN and its negotiation power with content producers, as well as the customer base of an ISP. This is a new paradigm for revenue sharing between the two parties. We develop protocols and systems to enable the collaboration between ISPs and CDNs without the need to reveal sensitive operational information that lead to a win-win situation. By jointly optimizing the network and hosting infrastructure, end-user experience is also expected to be improved.


Provider-aided Distance Information System (PaDIS)

To improve end-user assignment to servers, especially from a performance perspective, we propose and deploy a Provider-aided Distance Information System (PaDIS). PaDIS is a novel system operated by an ISP that allows them to influence the assignment of users to servers by utilizing information available only to ISPs (network conditions and end-user location) as well as by exploiting server diversity exposed by the CDN server selection process. PaDIS ranks any user and available CDN server pair based on current network conditions and accurate network locations. It then returns back the most preferable servers to the end-user regarding a performance metric chosen by the ISP. As a result, PaDIS utilizes the deployed CDN server infrastructure to improve end-user performance. To evaluate the performance improvements we have implemented and deployed a PaDIS prototype. Our field test results show that significant improvements in download time up to a factor of four for content offered by popular content distributors can be achieved when PaDIS is used.

Content-aware Traffic Engineering (CaTE)

We argue that CDNs and ISPs can jointly take advantage of the massively deployed distributed infrastructures to improve their operation and end-user performance. We show that it is sufficient for CDNs and ISPs to coordinate only in server selection, not routing, in order to perform traffic engineering. To this end, we propose Content-aware Traffic Engineering (CaTE), which dynamically adapts server selection for content hosted by CDNs using ISP recommendations on small time scales. CaTE relies on the observation that by selecting an appropriate server among those available to deliver the content, the path of the traffic in the network can be influenced in a desired way. We present the design and implementation of a prototype to realize CaTE, and show how CDNs and ISPs can jointly take advantage of the already deployed, path diversity, and ISP detailed view of the network status without revealing sensitive operational information. By relying on traces from a large European ISP and the largest commercial CDN, we show that CaTE allows CDNs to enhance the end-user experience while enabling an ISP to achieve several traffic engineering goals.

Network Platform as a Service (NetPaaS)

Recently, major CDNs have formed strategic alliances with large ISPs to provide content delivery network solutions. Such alliances show the natural evolution of content delivery today driven by the need to address scalability issues and to take advantage of new technology and business opportunities. Moreover, major ISPs have announced the deployment of generic appliances (microdatacenters) within their networks to support Network Function Virtualization. We revisit the design and operating space of ISP-CDN collaboration in light of recent ISP and CDN alliances and advances in network function virtualization. We identify two key enablers for supporting collaboration and improving content delivery performance: informed end-user to server assignment and in-network server allocation. We report on our experience developing and evaluating a prototype system, NetPaaS, that materializes them. Relying on traces from the largest commercial CDN and a large tier-1 ISP, we show that NetPaaS is able to increase CDN capacity on-demand, enable coordination, reduce download time, and achieve multiple traffic engineering goals leading to a win-win situation for both ISP and CDN.