FAQ

OCEAN has identified shortcomings and issues with the current CDN delivery model over Internet, especially in the so-called “CDN last mile”. Today, the delivery of content from the last CDN node to the end-use is done over best effort only. With the expected increase in (mainly video) traffic and associated bit rates, there is a clear interest in extending the current “Global CDNs” with “telco-CDNs” to offer quality content delivery to end-users. Global CDNs and telco CDNs differ in the following:

• Telco CDNs are deployed more or less deeply in the telco network, are located closer to the users than Global CDNs, and therefore potentially have a number of advantages in terms of user QoE experience ;

• Telco CDNs are localized, and would therefore possibly collaborate with Global CDNs for the delivery of global contents on a worldwide Internet scale ;

No assumption  is made at this stage about who owns each telco CDN. A telco CDN is not necessarily operated by the telco controlling the underlying telco network. It might be a player different from the telco. A player owning Global CDN could also own telco-CDNs or a global infrastructure composed of Global CDNs and telco-CDNs.

This implies that traditional Global CDNs and telco CDNs are likely to evolve to a complementary role whereby - generally speaking - traditional Global CDNs provide efficient global distribution and telco CDNs take care of efficient local delivery to end-users.  Note that does not exclude the direct delivery of content by telco CDNs or by Global CDNs for specific cases and specific types of content.

Therefore, any content internetworking architecture or mechanism should take into account such complementarities, and the somewhat hierarchical relation between both types of CDNs.

The project does not only address technical issues but also business models. Several models seem possible, including the end-user paying more for QoS delivery, but also a model in which the global-CDN shares revenue with the telco-CDN to guarantee quality delivery to the content provider seems feasible.

It seems logical that existing Global CDN providers would like to avoid any revenue sharing, hence oppose to any open interfaces to a telco-CDN. However, when the benefits are compelling, and traffic increasing significantly, we believe that Global-CDN players will see the win-win situation of the proposed architecture in the mid and longer term.

Will OCEAN bring transport costs down to a significant extent?

OCEAN is not only investigating how to reduce transport costs with deeper caching and adaptive delivery techniques, but will also allow better quality delivery to end-users. Therefore one should try to compare future transport costs with the current model and with the OCEAN model. At first level of abstraction it is seems highly probable that a model in which the network can adaptively react to changing demand patterns with an objective to maximize the QoE, will serve the user better than a model in which all clients are in competition trying to get as much bandwidth as they can.

Can OCEAN improve scalability, QoS and QoE?

Several techniques investigated in the OCEAN project have a very high potential in improving scalability, QoS and QoE. Deploying caches nearer to the end-user allows saving on required bandwidth upstream from that cache. This bandwidth can be used to deliver more (other) content, or content in a higher quality (bit rate). In addition, the load on delivery servers will decrease as the servers themselves become more distributed over the network. Adaptive (SVC-based) video delivery can gracefully decrease the required bit rate for specific video streams without heavily impacting the QoE for the end-user. In case of (near) congestion this results in an overall improvement of the QoE.

Has OCEAN any impact on mobile/wireless internet delivery?

OCEAN has an explicit focus on wireline high quality video delivery over the Internet, but we believe that several techniques and mechanisms will also be applicable to mobile video delivery (possibly with some mobile-specific modifications).

Impact of OCEAN on home caching (PVRs), DVD vs VoD?

DVD rental is expected to decline in favor of VoD. In addition, PVR will be replaced more and more by catch-up TV. Both VoD and catch-up TV will benefit from the OCEAN architecture and mechanisms.

Impact of OCEAN on live television and radio stations?

A better Internet delivery architecture will provide many benefits for future live television as well. On one hand, live television will be easier accessible on the Internet as alternative delivery channel (but no explicit goal of OCEAN to investigate). On the other hand, the addition of the Internet-based delivery channel can be used to offer new experiences in which live television is augmented with additional (personalized) content.

Impact of OCEAN on hybrid broadcast/broadband  (HBB, HbbTV, Canvas) projects, whereby the internet path can provide browser based services, catchup TV, to Internet connected TV devices and set-top boxes?

OCEAN concentrates on audiovisual content delivery over Internet only. The focus of OCEAN is to improve the Internet path. It aims at enhancing any content delivery systems relying on that Internet path, including the hybrid broadcast/broadband  projects such as HBB, HbbTV and Canvas.

How OCEAN addresses the question of geolocation (copyright) and analytics?

The level of geolocation provided today by global CDNs will remain as OCEAN aims at extending these global CDNs. If finer grained geolocation is required, a telco-CDN definitely would be ideally placed to provide that info.

Analytics will be part of the interface between the telco-CDN and the global-CDN. (Details are to be defined)

How OCEAN relates to the network neutrality?

OCEAN aims to impact on the evolution of the content distribution architectures over Internet, in line with the regulation context. But OCEAN does not intend to take any direct initiative regarding network neutrality.