Will accelerating Internet traffic growth produce a bandwidth famine? Is congestion slowing down your Internet experience?

Posted on July 15th, 2008 in Viodi View by Alan Weissberger
Yesterday, I read a very thought provoking article in Broadband Properties magazine (I have a print sub).  It stimulated my thinking on Internet bandwidth growth and network congestion. I have recently been keenly aware of how long it is taking some web pages to load and I can only attribute that to Internet congestion (probably caused by all the video traffic people are downloading, streaming).    I’ve also noticed some recent hang ups in mlb.com 1.2Mb/sec video streaming.
 
So here’s the article in question and a few more on this topic.
 
Responding to the Exaflood:
 
An Internet Innovation Alliance panel in New York last month noted that network growth nationally is about 19 percent a year – but that (Internet) traffic is growing at 40 percent annually, or more. According to Deloitte Telecommunications Predictions, Internet traffic doubles every 12 to 15 months. The Associated Press reports that YouTube users download more than 100 million videos a day. Professor Andrew Odlyzko of the University of Minnesota (x-AT&T) calculates that if YouTube traffic were converted to HDTV format, the downloads would equal all other traffic traveling on the Internet in 2007. Current Netflix traffic, Odlyzko says, would amount to 5.6 exabytes per year if Netflix videos were delivered online in high-definition Bluray format – about 10 percent of current. IP traffic in the US.
 
The San Francisco Chronicle calculates that by 2010, only 20 typical homes will generate as much network traffic as the entire Internet produced in 1995.
 
To warn of the coming Exaflood and to get some idea of how high the Internet "waters" will rise, the Internet Innovation .Alliance (IIA) held a conference in New York last month. Odlyzko, who has long resisted being an alarmist, said Internet traffic is growing at 50 percent per year, outstripping the mere 19 percent annual growth in Internet infrastructure – storage, server farms, and transmission facilities. He said the results could be disastrous.

Indeed, there was consensus among the panelists that at the moment, the telecom industry is not making necessary investment in plant and equipment  despite business opportunities for companies that can exploit the new Exaflood demand for goods and services. 

 
Quote from an article in the same issue,  Broadband to the Home: Broadband America:
 
"One researcher recently reported that in December 2007 a record 10 billion videos were viewed online. The largest US broadband provider says consumer broadband traffic on its network has doubled in the last two years alone, and broadband customers are using 40 percent more bandwidth per year."
 
Expert predicts global bandwidth famine -Growing demand will outstrip supply
 
Market experts have warned that the world’s consumers are facing a " bandwidth famine." The Global Bandwidth Study, commissioned by photonics firm CIP Technologies, predicts that the demand for internet bandwidth will more than double in two years and grow by an "order of magnitude" in five years. This accelerating appetite will place excessive demands on current network architectures, according to report author David Payne of the Institute of Advanced Telecommunications at Swansea University.

http://www.vnunet.com/vnunet/news/2221540/expert-predicts-bandwidth-famine

http://www.itpro.co.uk/604527/study-predicts-serious-bandwidth-shortages

Global bandwidth to double in two years: 

Global bandwidth demand will double in the next two years and usage could be 40 to 100 times what it is currently by 2018, according to a study.   The study, commissioned by CIP technologies, says demand will be over 160 terabits per second in 2010, which is more than the total demand from 1998 to 2008.  

Demand has risen due to the popularity of video sharing sites like YouTube, which some have claimed used as much bandwidth in 2007 as the entire internet in 2000.   ‘Networks are now being used in a way that few people foresaw, for example early take-up of personalised video, rather than broadcast television, dominating internet video services,’ said the study’s author, David Payne of the Institute of Advanced Telecommunications at Swansea University.

http://www.bcs.org/server.php?show=conWebDoc.20186

Opinion: Not enough fibre

According to experts the world will soon run out of Internet bandwidth unless we install oodles more fibre optic cable. This shortage was discovered by a company which makes…..fibre optic cables.

 http://www.crn.com.au/News/80396,opinion-not-enough-fibre.aspx

 Contrary view point article:  Don’t Fear The Bandwidth Apocalypse

A good rule of thumb: when someone claims the Internet is facing bandwidth armageddon, it’s usually because they’re in the business of designing and selling traffic shaping hardware, trying to justify new and frequently unjustifiable broadband pricing models, or trying to scare politicians into doing what they want. The guys actually working in the network operation centers will generally tell you that congestion can almost always be handled with smart design and capacity upgrades.

Last week the National Cable and Telecom Association (NCTA) was busy trying to lobby the FCC, which has been investigating exactly what sort of network management should be allowed, and how it should be disclosed to consumers. The NCTA argued that the use of deep packet inspection hardware was absolutely necessary on cable networks. Without such technology (the likes of which is being used to throttle Comcast P2P users), the NCTA claims that the Internet would all but collapse.
 
 
 
A recently updated Cisco study on Internet bandwidth measurements:

In Closing: Thoughts to Ponder:

So what’s your opinion on the current state and future growth vs capacity of Internet bandwidth?  What can network operators/ ISPs due to alleviate congestion caused by massive video downloads/uploads, streaming, and peer to peer traffic?   One IEEE ComSoc-SCV Discussion Group member wrote:

"More interesting than the data and high level inferences  in this
 paper, is the question of the architectural and technical approaches to
 meeting the need described here. I would guess (grossly simplifying) that
 intelligent caching (including P2P at the edge) to take advantage of the
 highly asymmetrical bandwidth patterns and "trickle feeding", together
 with the availability of really cheap storage (that’s at Moore’s Law/
 Moore’s Law+ growth rates) will be the philosophical approach for a lot
 of the video traffic. "

Note that the FCC has proposed to discipline Comcast for slowing down, meetering, or blocking their customers peer to peer traffic from Bit Torrent.  That’s a big vote for net neutrality.

http://deseretnews.com/dn/view/0,5143,700242562,00.html

What else should be done?  Do the network operators need to be more agressive in their FTTH/ FTTP deployments -like Verizon (FiOS) and some independent telcos?

Hopefully, some of what you read here will strike a chord and you’ll reply: 

alan@viodi.com

——————————————————————————————————————————

Addendum:  Network Congestion prompts MLB.COM to Prevent Customers from watching live games- Aug 8, 2008

On August 6th, I received bogus blackout messages for each live mlb.com game I tried to watch. This was independent of the streaming speed (I’m a premium subscriber that usually watches at 1.2Mb/sec). When I tried to access a different game, I received a page with this message: "You’ve reached this page in error." When I entered my zip code=95050, I got a messsage back indicating that I was blacked out for SF and Oakland. But I was not trying to watch either of those teams!

When I called tech support, I was told there was a mismatch between my IP address (in San Jose, CA) and the games I was trying to watch. That seems to be a huge software error in the MLB.com server. Then the tech said that "during periods of network congestion users sometimes get bogus blackout messages."  Finally, he had me remove and re-install a new version of the Silverlite media player. That seemed to fix the problem, but why? Is there a software bug in the previous version of Silverlite and if so, will the new version fix the problem (e.g. by allowing for a deeper playback buffer?   I just checked the mlbsupport forum and another user complained about the exact same false blackout problem.

http://www.mlbsupport.com/forum/viewtopic.php?f=5&t=1013&sid=7bc8c8647654ee3fd6e66ad926f6f38b

Popularity: 27% [?]

Comment?