“Toward Clean-Slate Future Internet Design:
SMART – An Optical Networking Perspective”
S.Q. Zheng, UT Dallas
Abstract
The current Internet faces many challenges. There are two paths of addressing these challenges: (1) Introduce point solutions (patches) of narrow scope to address vulnerabilities and new opportunities as they occur. (2) Create a new Internet architecture (clean-slate design) that addresses the challenges on the horizon. The second path is being seriously considered.
We introduce a new optical network architecture called the Scalable Multi-Access Reconfigurable Transport (SMART) network and discuss its advantages over the current Internet. SMART is based on the hypernetwork theory and applications. A key innovation of the hypernetwork is its capacity to provide connectivity by establishing hyperchannels, each of which may consist of two or more nodes, in the network. Within a hyperchannel, connectivity of different nodes can be achieved by multiple access schemes. Moreover, multiple hyperchannels can be used to establish a hyperpath. SMART network does not insist on all-optical communication (although that is a possible option). It is a hybrid approach that combines the best features of optical transmission and electronic processing, allowing optimized control and management. It reduces switching, forwarding and related packet processing to a minimum, solving the bottleneck of packet switching. SMART solves the scalability problem by utilizing bandwidth in arbitrary granularity. SMART provides service-oriented, application-specific connections of guaranteed performance by emulating circuit-switched all-optical communication without losing scalability. It supports layer-1 private virtual networks and various overlay networks, and it allows additional security and reliability features to be built on top of network virtualization. SMART is feasible because it can be materialized by integrating existing mature optical and electronic technologies.
Bio
S.Q. Zheng received his Ph.D. in electrical and computer engineering from the University of California, Santa Barbara in 1987. He is a professor of computer science, computer engineering and telecommunications engineering at UT Dallas. His research interests include algorithm design and analysis, circuits and systems, parallel and distributed processing, and telecommunications and networks, and he has published extensively in these areas.
