The title of his lecture, which was given in the Massachusetts Room of the Mullins Center to a standing room audience, was "Design and Analysis of a Content-oriented Internet."
Professor Kurose began his lecture by saying that he is "A Teacher" and provided us with an overview of the major developments in communications from the telephone to packet switching to the Internet. He emphasized that the Internet is a Network of Networks and, since I founded and direct the Supernetwork Center at UMass, I liked hearing this very much. Supernetworks are networks of networks and our research at the center has advanced both methodologies and the modeling and solution of applications ranging from multimodal transportation networks to complex supply chains and to now envisioning a Future Internet Architecture (FIA), which is funded by the National Science Foundation (NSF), and which we are working on with Co-PIs at multiple universities. Our FIA is known as ChoiceNet.
Professor Kurose emphasized that "Content is King" and stated that Cisco is predicting that, by 2017, 80-90% of the Internet traffic will be for videos, TV streaming, peer to peer, etc. He noted that consumers don't care where the content comes from as long as we get it.
Everything used to be wired but now there are more wireless devices than wired ones connected to the Internet. There are also millions of access networks - for example, at UMass we would first be connected to that network.
He also spoke about the challenges that Netflix faces and that researchers in the Computer Science Department are working on. Some of my students and I had had a preview of this topic just the day before in my Management Science doctoral seminar on Variational Inequalities, Networks, and Game Theory, since a Computer Science professor and one of his doctoral students have been sitting in this course of mine and it has been terrific to have them. We had heard the day before about how Netflix is unsure of where to position movies (store them) and makes multiple copies in multiple locations - this may reduce the speed of delivery and also add to network congestion plus storage costs - topics I have spoken about.
Afterwards there were great questions, even from our Chancellor - not surprising since he is a physicist!
I commented that the Netflix problem reminded me of a problem in humanitarian logistics (another course that I am teaching this semester) in terms of the positioninng of supplies and also mentioned our FIA project ChoiceNet, which will allow consumers to select not only content but the mode of delivery which could include less congested paths with consumers willing to pay for such enhanced services as we say "pay with their wallets."
Indeed, in our ChoiceNet project, which we already have multiple papers on, we draw comparisons with supply chain networks and freight delivery systems and recognize the importance of network economics, game theory, and operations research. Professor Kurose, after my comments, did note that I was in OR and mentioned "facility location." I have a good idea as to how to model the Netflix problem and solve it. He also mentioned that 2 out of the 5 NSF FIA [projects had UMass Amherst contributors, which is wonderful!
After his fabulous talk - clear why he has received multiple teaching awards research recognitions, I took the photo below of him with my doctoral students.
Professor Kurose has been at UMass Amherst for 30 years and over the next 30 we can continue to ask him many questions that his talk inspired.
I remember, back in 2005, sharing the stage with him at the Bowker Auditorium, when both of us were amongst the 12 faculty being recognized with a new award, which is now given annually - the Award for Outstanding Accomplishments in Research and Creative Activity.
Great to have the opportunity for my doctoral students in Management Science to meet him yesterday!
Posts
