Can Colleges Tame The Bandwidth Monster?

Imagine that you're starting college today. Your dorm room -- excuse me, residence hall suite -- has a cable TV jack and an Ethernet jack in the wall. But you didn't think to bring a TV or a device that even has an Ethernet connection. The college has wireless access, but it's far too slow to stream your regular TV programming to your tablet computer. Your roommate has a laptop that can use that Ethernet jack, but he says it's even slower than your wireless connection. You try your cell phone and you get two bars inside the room, at best. You need to download your textbooks for the semester, but more than 2,000 other students are doing the same thing at the same time. You want to call home to complain, so you have to run outside to get a cell signal because the wireless network isn't fast enough to Skype.

You begin to panic, and then get that sick feeling in the pit of your stomach. You have been (dun dun dun) disconnected from the world!

As a higher education CIO, I often try to put myself in the place of students, and this is the most dramatic scenario I've come up with so far. Unfortunately, for many college students this scenario isn't far off the mark. Welcome to college life in our digital society.

Each year, I'm faced with the continual challenge of feeding the Internet bandwidth monster. It's a strange industry when higher ed CIOs are elated about the next big thing in technology while also terrified by the effect it may have on our already congested networks and Internet connections.

[ Are printed textbooks on the way out? Read E-Textbook Pilot Puts College Books In Cloud. ]

Maryland-based State Educational Technology Directors Association recommends that a national standard for bandwidth be set for K-12: 1 Gbps per 1,000 students and faculty/staff members for Internet connections and 10 Gbps for LAN connections, all by the 2017-18 school year. If this is to be the standard for primary and secondary schools, what might the standards be for a small college or university? And what about our major research universities? 100 Gbps to the desktop? Terabit backbones? Higher? The nationwide costs could be staggering for the necessary wiring and hardware upgrades. Hang on -- I think I need to lie down a minute.

How Long Can We Feed The Monster?

Today, adequate Internet access is tied so directly to academic scholarship and research that we no longer ask ourselves whether adding more bandwidth is really worth the trouble and expense. For work or play, fat pipes are now just part of student expectations, like electric lights … and a campus coffee bar.

I've tried to keep up since the early 1990s with all sorts of network solutions. First, we threw more bandwidth at the problem. Then we made "creative use" of filtering tables, then packet shapers, now bandwidth equalizers. But we never solved the problem for long as the newest "disruptive" technology clogs our pipes.

With applications such as Skype, Netflix and Xbox, our bandwidth situation was bad enough, but now with cable and satellite content providers piling into the mix with their mobile apps, how do we keep up with demand? Even more disruption is coming with new augmented-reality applications and devices, so gear up!

Can We Tame The Monster?

If the monster is demand, my answer to the above question is, "No way." But I'm hopeful that technology innovations will keep us a step ahead of growing demand.

WAN research through the Internet2 organization lights the way to the future. Rod Wilson, senior director of external research at Ciena Corp., wrote in a 2011 blog that 100-Gbps long-haul transmission and switching have solved myriad "well-publicized problems" for providers of communications services to research institutions, "including bringing relief to saturated fiber plants suffering from nearly full links." Commodity ISPs are just starting to deploy these more-efficient, higher-speed technologies, promising higher-bandwidth services at lower cost.

So much for the wide area. But how do we attack the head of the bandwidth monster, the constantly growing need for faster wired and wireless connections inside our college campuses?

My good friend Jimmy Ray Purser, a network engineer with Cisco, says the major problem today with most campus networks isn't the amount of bandwidth but the optimization of that bandwidth. "About 40% of your traffic is overhead chatter," Purser says. "The trick is to have a good VLAN design and newer switches to capture as much benefit as possible with the latest compression algorithms. But the best features of bandwidth optimization tools are not being used in many colleges and universities today."

Purser maintains that OpenFlow could be the answer to many of our campus bandwidth problems. OpenFlow is a vendor-agnostic standard that allows network switches to be controlled by a central network operating system. With OpenFlow, packet-forwarding instructions set up in the NOS make more efficient use of processing power and paths. This technology is now being tested in switch-to-server and switch-to-cloud (cloud bursting) applications, allowing for the next level of resource optimization and customization across the entire hardware infrastructure.

Higher ed IT organizations will still continue to deal with the challenges we have come to know and hate: file-sharing software, illegal music/video downloading, gaming and streaming video services. But the solace is that OpenFlow and technologies like it will provide a leap forward in campus network bandwidth and management over the next five years.

Will costs come down? I'm pessimistic. Where there's pent-up demand, there's always a higher price to be paid, even if carrier competition increases over the coming years. The key to the future is balancing bandwidth with new technology to slow the monster and keep it from swallowing us whole.

Additional Questions

I continue to have more questions on this topic than answers, and I'd like to get your ideas. Over time, will LTE and next-generation cellular transport technologies ease the strain on campus networks? Will bandwidth provision continue to be the responsibility of the institution only, or is it fair to ask students to augment college-supplied Internet service with their own data services? How would this shift from private campus networks to public cellular networks affect security?

Please weigh in with a comment below.

Can data analysis keep students on track and improve college retention rates? Also in the premiere all-digital Analytics' Big Test issue of InformationWeek Education: Higher education is just as prone to tech-based disruption as other industries. (Free with registration.)