COMIT: Content Management at Internet Scale

Vision

The Internet is currently passively pushing bits between end-host machines, be it servers, fixed or mobile user devices, or sensors. The network does not "understand" what is being transferred, i.e., it is not content-aware. This agnostic mode of operation affects several of its key functionalities, for example, efficient content distribution and content-aware traffic engineering. As a result, the network is not able to cope well with the exponentially increasing amounts of multimedia content access which constitute the major mode of use in recent years. Fixed and mobile network providers keep continuously upgrading their infrastructures but the situation has become unsustainable due to their eroding profit margins. There is an urgent need to rethink traffic management under the umbrella of active content management, rather than passive content transfer, allowing ISPs to control traffic better and achieve a sustainable model for the long-term evolution of their networks. New approaches that maximise traffic localisation are essential for long-term global network sustainability.

Recent research in the general area of Information-Centric Networks (ICN) has revealed the benefits brought by a content-focused mode of operation. Little attention, however, has been payed on the deployment incentives and the migration path from today's IP-dominated Internet architecture.

The COMIT project takes a closer look at the deployment challenges of Information-Centric Networks, focusing primarily on preserving the current Internet infrastructure, but at the same time taking advantage of the benefits of information-centricity and location-independence.

The primary objectives of COMIT are:

  • Support the incremental and smooth deployment of ICN but at the same time benefit directly from all the important features of content-aware or Information-Centric Networking.
  • Allow ISPs to migrate to this "hybrid" ICN operation but at the same time maintain full backwards compatibility with ISPs that do not migrate and operate those ICN features.
  • Give the opportunity to ISPs to do content-aware traffic engineering in their own network, deploying their protocols of choice, and making best use of their resources.



The COMIT Team @ UCL.
Funded by the EPSRC