Toronto Networking Seminar

Organized by Department of Computer Science and Department of Electrical and Computer Engineering, University of Toronto


 

Network Services Scheduling for Emerging Applications

 

Nasir Ghani 

Advanced Cyber-Infrastructure Lab

ECE Department, University of New Mexico, Albuquerque, NM, USA

http://ece.unm.edu/~nghani, http://ece.unm.edu/acl  

 

 

Date: March 18, 3pm

Room: BA B024 (Bahen Center Basement)


Abstract:

The last decade has seen many advances in next-generation networking technologies.  For example, ubiquitous IP and Ethernet networks (Layers 2, 3) have evolved to multi-gigabit speeds with full quality of service (QoS) provisions, as enabled by gains in high-speed electronics technologies.  Concurrently, developments in optical wavelength division multiplexing (WDM) sub-systems have revolutionized the fiber-optic layer (Layer 1), delivering flexible “wavelength” circuit connectivity with terabits/fiber yields.  As these technologies have matured, many scientific research organizations have actively deployed wireline networking infrastructures to support expanded distributed computing needs.  For example, many “e-science” applications already need massive dataset transfers, i.e., petabytes-exabytes range, placing huge burdens on network resource provisioning.  Given these trends, it is very plausible that even the most scalable backbones may not be able to handle all user requests in an “on-demand” manner.  As a result, the concept of advance reservation (AR) of connections is becoming an increasingly critical concern.  Namely, the ability to reserve connections at future time instants, i.e., network scheduling, allows operators to stagger demands and improve resource assignments and utilization.  Moreover, AR services also have broad relevance to data-center management and other commercial applications.  In light of the above, various studies have looked at this area, both at the IP/Ethernet and optical DWDM network layers.  Along these lines, this talk will survey this exciting new field and outline several new research directions.  In particular, refined AR scheduling strategies will be presented to help lower request blocking rates and improve resource efficiencies.  Furthermore, novel network control plane extensions will also be detailed to help translate theoretical AR algorithms into practical real-world settings, i.e., distributed operation, multiple domains.

 

Bio:

Dr. Nasir Ghani is an Associate Professor in the Department of Electrical and Computer Engineering at the University of New Mexico, where he is actively involved in a range of research and teaching activities.  His current interests include network cyber-infrastructure design, survivability, services/applications, and knowledge-based systems. He has published over 40 journal and 80 conference papers, several book chapters, and has also co-authored various industry standards contributions.  Overall, his work has been funded by several US government agencies (including the National Science Foundation and Department of Energy) as well as some industry sponsors.  In particular, he received the NSF CAREER award in 2005 to support his work in the area of multi-layer network design.  Prior to joining academia, Dr. Ghani spent over 8 years working in industry and held key technical positions at several large companies (Nokia, Motorola, IBM) as well as some start-up organizations (Sorrento Networks, Array Systems Computing).  In addition, he has actively been involved in a range of outreach and technical community service roles. For example, he has served as chair of the IEEE ComSoc Technical Committee on High Speed Networks (TCHSN) from 2008-2010 and has also been a symposium co-chair for IEEE Globecom (2006, 2010) and IEEE ICC (2006, 2011).  He has also built and established a successful high-speed networking workshop series for the flagship IEEE INFOCOM conference and served as a panelist for numerous NSF, DOE, and international panels.  Dr. Ghani is an associate editor of the IEEE Communications Letters journal and has also guest-edited special issues of the IEEE Network, IEEE Communications Magazine, and Cluster Computing journals.  He is a Senior Member of the IEEE and has also been faculty advisor for the Etta Kappa Nu honor society. He received his Bachelors degree in computer engineering from the University of Waterloo, Canada, his Masters degree in electrical engineering from McMaster University, Canada, and his Ph.D. degree in computer engineering from the University of Waterloo, Canada.

 

 

Host of Talk:

Yashar Ganjali [yganjali@cs.toronto.edu]