Open Network Interfaces for Carrier Networks

A. Panda, M. McCauley, A. Tootoonchian, J. Sherry, T. Koponen, S. Ratnasamy, S. Shenker
Appears in: 
CCR January 2016

With the increasing prevalence of middleboxes, networks today are capable of doing far more than merely delivering packets. In fact, to realize their full potential for both supporting innovation and generating revenue, we should think of carrier networks as servicedelivery platforms. This requires providing open interfaces that allow third-parties to leverage carrier-network infrastructures in building global-scale services. In this position paper, we take the first steps towards making this vision concrete by identifying a few such interfaces that are both simple-to-support and safe-to-deploy (for the carrier) while being flexibly useful (for third-parties).

Public Review By: 
David Choffnes

There is an active discussion about whether carriers should simply forward packets on behalf of end-to-end services, or whether they should leverage in-network devices -- middleboxes -- to improve the user experience. One thing is not under debate: middleboxes are pervasive in today's networks and they are unlikely to go anywhere soon. However, today's middlebox services are deployed in an ad-hoc manner, and they are owned, operated, and controlled by carriers with little or no input from the applications they interpose on. This paper argues that instead of such a closed approach to implementing in-network functionality, carriers should sell access to network functionality as a service with open interfaces, much like Amazon sells access to their EC2 infrastructure. The authors sketch how to enable this vision, called Network Service Support, or NSS. They discuss the entities involved, and the interfaces they can use to access in-network compute and storage resources. While they do not focus on any particular instantiation or implementation of a service built on open network interfaces, they do describe several expected systems that incorporate caching and video multicasting, among others. The reviewers were generally positive about this position paper, but raised several issues and considerations. Perhaps the biggest issue is one of incentives: why would a carrier sell access to their infrastructure if it reduces their profit from transit charges or competing technologies (e.g., video streaming)? Another common concern is that without Internet-wide standards for these open network interfaces, it might be prohibitively expensive for an application provider to take advantage of disparate NSS deployments. Regardless of the deployability of the approach, the reviewers appreciated the vision and its potential. As we move to an Internet where content is increasingly encrypted end-to-end, the NSS model may become the only way for carriers to support traditional middlebox services -- by bringing the end to the middle.