GIGABIT INTERNET Gig.U's goal is to accelerate the deployment of next-generation networks in the U.S. by encouraging the development of new applications and services that can make use of ultrafast data transfer rates.
Image: COURTESY OF DEBSTREASURES, VIA ISTOCKPHOTO.COM
-
The Best Science Writing Online 2012
Showcasing more than fifty of the most provocative, original, and significant online essays from 2011, The Best Science Writing Online 2012 will change the way...
Read More »
In the not-too-distant future broadband speeds will be measured in gigabits per second rather than megabits per second, the former being 1,000 times faster than the latter. Such blazing fast data transmission will vastly improve the quality of streaming high-definition video, playing online video games, participating in video conferences and using voice over IP, all of which struggle with latency at today's average data transfer rates, which range from less than one megabit per second (Mbps) to 10 Mbps (pdf). The sticking point over gigabit-per-second broadband: who will pay for it?
Telecommunications companies, still stinging from the financial beating they took a decade ago after hastily building up capacity for Internet companies that soon went out of business, have been leery ever since of investing in infrastructure unless they are certain there is a demand for it. Most customers, many of them still exploring the wonders of YouTube and for the most part content to simply use e-mail and social networks, are not demanding, nor are they willing to pay a premium for, service that moves information at 1 billion bits per second.
The exception lies at the seat of learning—universities and research institutes that can find a way to use any extra bit of speed that their ISPs can provide. With the federal government in no position at the moment to invest heavily in the National Broadband Plan introduced by the Federal Communications Commission (FCC) last year, more than 30 universities and counting have taken the matter into their own hands, forming the University Community Next Generation Innovation Project, more commonly referred to as Gig.U. Members include schools across the country—from the University of Alaska down to the University of Florida at Gainesville.
Gig.U's goal is to accelerate the deployment of next-generation networks in the U.S. by encouraging researchers—students and professors alike—to develop new applications and services that can make use of ultrafast data transfer rates. Imagine downloading two HD movies in less than a minute or expanding the use of video conferencing and Webcasting without worrying about latency issues that today pixilate images and freeze streaming video.
The most interesting uses are yet to come, according to Gig.U organizers, who want to extend gigabit-speed networks beyond campuses and into the surrounding college communities where students and professors live. College communities make sense as test beds for gigabit networks because they include highly concentrated population of heavy Internet users as well as institutions already connected to Internet2, National LambdaRail (NLR) and other high-speed Internet backbones, says Gig.U Program Director Elise Kohn, a former policy advisor in the FCC's Wireline Competition Bureau. "If you look at a lot of our international competitors in the research community, we think our researchers will be left behind without gigabit speeds," she says.
The gigabit networks will vary from site to site, depending on the approach that different ISPs propose to meet the differing needs of Gig.U members. "All of our members are focused on next-generation networks, although some will need more than a gigabit while others will need less," Kohn says. Gig.U is holding a request for information (RFI) from September to November to solicit ideas from the local service providers for building out new networks, which would ultimately be funded by Gig.U members as well as any non-profits and private-sector companies interested in the project.
Although there is no formal connection between Google and Gig.U, Kohn says she would welcome Google's involvement with the project, which is being led by Blair Levin, who headed development of the National Broadband Plan and has stressed the importance of test beds for high-speed connectivity.
In fact, Gig.U was inspired in part by the Google Community Fiber Initiative, a project the company announced in February 2010 to set up ultra–high-speed broadband networks in a small number of trial locations throughout the U.S., says Lev Gonick, chief information officer at Gig.U member Case Western Reserve University in Cleveland. Case Western has already created a network in line with Gig.U's goals called the Case Connection Zone, which connects more than 100 residences near the university via a gigabit fiber network.
Google demonstrated that the demand is out there, Kohn says, adding, "Look at how the communities worked to get Google to choose them as part of their program." Within one month of announcing its Community Fiber Initiative, the company had received more then 1,100 responses from communities throughout the U.S. requesting to become test sites for Google's new network. Google plans to announce the test site or sites by the end of the year. Many of these submissions included homemade videos pleading their case. Sarasota, Fla., Mayor Dick Clapp even donned scuba gear and climbed into a shark tank to show his commitment (video). Less clear is what these different communities plan to do with a gigabit network if they are chosen as a test site.
This is not the case at universities, many of which now tout their networks to attract professors and students working on ambitious research projects. Gig.U officials are hoping that at least some of these projects will spur interest in gigabit networks and seed demand for higher bandwidth nationwide.
See what we're tweeting about
12 Comments
Add CommentSad to see the old error of assuming latency and data transfer rates are always linked being used here. If you have naturally high latency (e.g. a satellite link), then no amount of increase in your data transfer rate is going to reduce it. Having said that, I'm always happy to see faster networks being rolled out.
Reply | Report Abuse | Link to thisWho will pay for it? Where is the demand? Aside from full movie downloads, is there any real requirement for such high transfer rates?
Reply | Report Abuse | Link to thisThe issue is primarily that communications is only as fast as its slowest link - the entire nationwide communications infrastructure would require nearly a complete replacement to ensure that point to point transfers could meet the gigabit target rate.
If only a subset of internet communications customers demand the ability to download movies at gigabit rates, that could be achieved by their paying sufficient rates to finance gigabit links between themselves and local or regional movie file servers.
Like it or not, that capability could also be achieved nationally through the use of satellite links, which replace physical terrestrial links with 'virtual' microwave radio links - just like local cable providers who download video services from satellite broadcasts. However, movie downloads require not just a broadcast of TV channels to all subscribers but a specific point to point link between a file server and each subscriber request.
Frankly, I have no need for any services that require national (or global) gigabit communications and do not wish to pay the enormous infrastructure investments necessary to support those that do.
In fact, my bandwidth requirements would be significantly reduced if internet advertisers were prevented from sending unsolicited video advertisements on web pages that I am interested in.
If must watch video on your phone, please restrict your viewing to NFL football or broadcast movies, and pay for the speeds you require.
L'activité des sourires.
Reply | Report Abuse | Link to thisComme une
ombre dans
l'intensité d'un
matin je sens
le soleil qui
chante avec toi
le son des
sourires.
Francesco Sinibaldi
In the case of universities that are doing research, I think the high speed networks would greatly speed up networked number crunching at a cost much lower than investing in supercomputers. Sharing of data and shared processing of data would greatly benefit from the higher speeds. Remember that a lot of the research using data from Cern and other scientific instruments is done far from those instruments and getting the volumes of data to do that research as well as analyzing that data is timeconsuming.
Reply | Report Abuse | Link to thisTo try to justify it based on movie download speeds is foolish and is more likely an attempt to get the public to clamor for it. The public is only interested, for the most part, on the entertainment value they get off the internet, not in the original purpose of the internet which was to share information between scientists.
In Canada we have Canada's Advanced Research and Innovation Network (CANARIE) which hooks up our Universities here. The info on their website says "Currently, all wavelengths are 10Gbps unprotected point-to-point connections, which terminate on transport switches located in CANARIE Network Points of Presence (PoPs)."
Reply | Report Abuse | Link to thisDon't Universities in the US already have something similar?
With no interest from the telecoms, the universities think they can achieve gigabit speeds? I doubt that lack of demand is the problem, as faster speeds are always a sight for sore eyes, reducing costs and frustration for everyone. Providing last-mile optical fiber to residential housing is out of the question, but without it, there's no need to discuss anything faster than coaxial - 10Mb/s. Massively-parallel, terabit computers now cost less than US$3 or 4 million, making it an embarrassment to even mention shared computing at cocktail parties. But, most importantly, the silicon isn't capable of such switching speeds, even if bandwidth wasn't the limiting factor. And wireless can't handle email and voice in a densely populated area, let alone video. In short, old windbags living among the ivory towers of academia don't know anything worth knowing, and are hoping that by pretending that gigabit speeds are just around the corner, that some "greater fool" is going to throw funding their way.
Reply | Report Abuse | Link to thisAs RHoltslander points out, point to point gigabit connections are commonly used in business and I presume university environments where requirements justify the required cost, individually funded by each organization or institution.
Reply | Report Abuse | Link to thisThe implication of this article is that communications companies will implement the necessary infrastructure to provide public access to gigabit internet connections.
Foolish me, I merely ask what application the public demands can justify the incredible costs required for public gigabit internet connections? It all depends on who's paying - the general public is being sold on the ability to download movies 'instantaneously', but they are not being informed of any cost to do so.
Is there some other application that requires and justifies public gigabit internet connections?
This is old news,"Chattanooga Announces Only 1 Gigabit Broadband Service in U.S. for Residential and Business Customers September 13, 2010" you can download and upload at the same speeds.
Reply | Report Abuse | Link to thisBut it is VERY expensive, $349.99 a month
Well I will point out that there were plenty of people wondering why you'd ever want a vehicle that goes over 30 miles per hour when a horse was plenty fast...
Reply | Report Abuse | Link to thisAnd I recall lots of people wondering why you'd ever need a (gasp) 10 megabyte hard drive on a PC...
Of course there were those that wondered why individuals would ever need their own PC... those things were massively expensive and really only made sense for universities... and what would a house wife do with a home computer anyway? File away recipes?
The point is that the costs come down and the public finds a use for this stuff, then can't imagine the days when they didn't have the extra capacity.
Point to point gigabit connections are one thing. Modifying the infrastructure to support the bandwidth that translates to the local 1 gigabit connection is another. The bandwidth to the servers supporting the internet would need to expand to support multiple 1 gigabit connections halfway around the world necessary to connect a number of universities in Europe and the Americas with a telescope in Australia as an example.
Reply | Report Abuse | Link to thisWe have 1 gigabit networks in most work places now but they are local. There are also 10 Gigabit and even 100 Gigabit networks as well which would also require upgrades in the backbone of the internet. Not something I need for cruising the internet currently but something that would help when I have to remote into a computer on the opposite coast of the US or somewhere else. As a consultant working from home, the slow connections I can get seriously impede my ability to resolve problems remotely.
So I guess the question should be, what do you use the internet for and would you yourself benefit from a higher speed backbone for the internet? I would benefit but then again, at what cost. Once the backbone is in place, I would need to upgrade my equipment at home as well. The current 1 gigabit netork interfaces would need to be swapped out for 10 or 100 gigabit interfaces. Maybe even faster processors and PC's/laptops. So there is a personal cost to being able to take advantage of a higher speed internet as well.
Aside from socio-economic-political considerations regarding a Gbps Internet, it would behoove SciAm to whet the readership’s collective palates by devoting, perhaps, 20% of their news-item “bandwidth” to the scientific and technological issues that must be addressed before such a next-generation network can become a reality. Examples may include such challenges as source development, multiplexing, modulation and detection techniques, routing and distribution, dynamic interconnects, extremely high bandwidth non-dispersive materials and structures, energy considerations, etc. Questions such as “What critical fundamental and practical roadblocks must be overcome?” may be just what the doctor orders to inspire curious minds to enter this exciting and relevant field of endeavor. Achieving a 1000-fold leap in bandwidth is not necessarily an exercise in simple scaling of existing design rules, but, most likely, will require entirely new technical approaches. Devoid of such discussions, SciAm may be misconstrued as a non-technical periodical that merely sweeps science and technology under the rug.
Reply | Report Abuse | Link to thisThe particularly cheap equipment that I use in my home network is already quite a bit faster than the speeds I get from the local cable internet provider and leaves the DSL alternative in the dust. Until we have meaningful high speed to the home the rest won't really matter.
Reply | Report Abuse | Link to this