OCARINA: Optimizations to Compel Adoption of RINA

RINA, the Recursive InterNetwork Architecture by John Day [1], is a novel “back to basics” type approach to networking that is fundamentally different from today's networks. It shows great potential in many aspects of networking, e.g. by simplifying management and providing better security. The recursive nature of RINA calls for radically different approaches to how networking is performed, including congestion control and routing. In OCARINA, we will develop new congestion control and routing/forwarding mechanisms for RINA. Today, in the Internet, congestion control is performed end to end, but by locating the control closer to the entity that is being controlled, RINA should allow us to achieve much better performance. Also, in the Internet, routing is very static; in OCARINA we will research new dynamic ways to perform routing and forwarding. Our new mechanisms will be implemented, tested and evaluated in RINA, and we will see how RINA can be gradually deployed in (over / under / alongside) the Internet.

TCP/IP vs. RINA (adopted from [2], which was originally inspired from Tomáš Podermanski)

About the project

In OCARINA, we team up with leading international researchers to address these challenges. We expect to be able to show that RINA, with our new mechanisms, is indeed a much better solution for the Internet than TCP/IP in terms of performance. A goal is also to move RINA closer to real world deployment and motivate its adoption.

Objectives

  1. Research & develop new congestion control mechanisms in RINA,
  2. Research & develop new dynamic routing and resiliency mechanisms in RINA,
  3. Deploy RINA as an overlay/underlay/alongside the Internet. 

Outcomes

  1. Scientific articles,
  2. Source code of developed mechanisms as policies in RINA.

Background

The Internet's end-to-end'ness of congestion control and its very static routing are inevitable by-products of its underlying architectural design (e.g., in the Internet, it is virtually impossible to let load-based routing scale without obtaining oscillations). With the Recursive InterNetwork Architecture “RINA”, the natural way of applying such algorithms is very different, with control executed much closer to where the problem is. It is, therefore, an ideal vehicle for investigating drastic changes to how congestion control and routing could be done, and it provides the framework that we will work in.

Partners

  • John Day, Boston University, USA
  • David Hutchison, Lancaster University, UK
  • Eduard Grasa, i2CAT, Spain
  • Miguel Ponce De Leon, TSSG, Ireland

References

[1] J. Day, Patterns in network architecture: a return to fundamentals. Prentice Hall, 2008.

[2] Vladimír Veselý, Marcel Marek, Tomáš Hykel, Ondřej Ryšavý, “Skip This Paper - RINASim: Your Recursive InterNetwork  Architecture Simulator”, in 2nd OMNeT++ Summit, 3-4 September 2015, Zurich, Switzerland. [online] https://summit.omnetpp.org/archive/2015/assets/pdf/OMNET-2015-13-Slides.pdf

Tags: RINA, Congestion Control, Routing
Published Dec. 19, 2016 12:17 PM - Last modified Feb. 16, 2017 12:25 PM