The Internet most people use is an antique.
I hadn’t thought of it that way until I was invited to participate in a bonus session on Broadband Planning & Big Data, during the #APA13 National Planning Conference in Chicago last month. The International Center for Advanced Internet Research (iCAIR) at Northwestern University focuses on developing advanced large-scale networks supporting data-intensive (primarily scientific research) applications, and associated public policy. They are all about broadband infrastructure.
The public Internet in use today has its roots in defense and academic networks developed in the 1960s, standardized with protocols in the 1980s. iCAIR Director Joe Mambretti discussed the wide variety of “Grand Challenges” advanced research topics the organization is taking on to help the grand old Internet keep up with the times. Besides scaling from a billion users to upwards of 8 billion connections in the “Internet of Things”, global networks face the need to migrate services to new protocols and enable rich multi-media services.
All this takes bandwidth. Lots and lots of bandwidth. Bandwidth the existing, antique Internet doesn’t currently have, and can’t deal with well when it does have room. Add to the technology challenge, an upended communications business model shifting from utilities delivering content in one direction to ubiquitous services based on large-scale distributed broadband facilities.
The next generation Internet architecture will not be a one-size-fits-all proposition. While open standards are vital for interconnection, a variety of hardware and protocols will permit flexible public and private networks to best meet individual needs. That, personally, makes me a bit nervous about maintaining the populist, open-access ethos of today’s Internet, but in some ways its analogous to the public highway system. Everybody can drive on the public roads, but sometimes we need toll roads and either way jet aircraft will get you there quicker if you’re willing to pay the price.
Science is driving the demand for high bandwidth networking, with the most advanced applications perhaps leading the consumer market by up to 30 years. The traditional Scientific Method of Theory + Experimentation is rapidly adding Big Data Modeling & Simulation to the mix. Big Data demands Big Bandwidth, which many research institutions are choosing to install themselves. Higher Ed is dropping thousands of miles of their own “Dark Fiber” instead of jacking into traditional communications networks. These private networks provide a backbone for distributed computational environments, supplementing (or surpassing) common Cloud Computing networks.
Sounds like Science Fiction
The biggest and coolest advanced applications read like a script from Star Trek. Ultra-high definition 4k video (8.3 megapixels) will be 4 times the resolution of HDTV… and is already becoming obsolete with development of 8k media (33 megapixels). CAVEs (Computer Aided Virtual Environment) offer a 3D modeling and simulation environment with applications from medical imaging to crisis response to really cool museum films.
But this Big Data broadband demand isn’t the future, it’s happening right now. New networks connect high energy physics experiments that helped confirm existance of the Higgs Boson. They connect radio telescope communities and enable large-scale climate models. And they enable the US Geological Survey to process data from the new Landsat 8 satellite, with each image generated a generous 1 GB compressed.
Higher capacity networking and the advanced applications that use them will likely come about through equal parts tinkering and gee-whiz discovery. As Mambretti pointed out to our group, each hair-thin fiber optic strand can carry many, many communications channels, and each channel can carry more data when we improve the transponders on each end of those fibers. In the future, we’ll look back at today’s Internet and wonder how we got along, the same as we might wonder about our horse-and-buggy ancestors. Balancing new protocols and new innovation, we can get there from here.
.
STrib