x
Optical/IP

Broadband Myths Debunked

Andrew Odlyzko, the researcher famous for debunking some popular claims of exploding Internet traffic growth, plans to stir up more controversy next week at two U.S. conferences.

The outspoken researcher will deliver keynote speeches at Light Reading’s two NGS 2004 conferences next week – one on each coast. The NGS 2004 conferences are being held in New York City next Monday, Sept 13, and in Burlingame, Calif. on Wednesday, Sept 15. Entry is free for service providers and enterprise users.

Odlyzko, director of the Digital Technology Center at the University of Minnesota, will pinpoint a number of “misleading myths” about broadband services, which he says are impeding reform and restructuring. These myths include:

  • Carriers can develop innovative new services. Odlyzko contends that users have created all of telecom’s killer apps, such as email, the World Wide Web, browsers, search engines, and peer-to-peer file sharing. The role of carriers should be to provide platforms for users to develop their own apps.

  • Streaming, real-time, multimedia traffic will dominate. Odlyzko backs the argument made by Morgenthaler’s Drew Lanza in a recent Light Reading column: File transer for local storage is going to rule (see Fiber's Sticky Wicket).

  • Content is king. Odlyzko contends that the dominant types of communications pertain to business and social activities – as they have for thousands of years.

  • Voice is passé. Nextel Communications Inc.'s (Nasdaq: NXTL) push-to-talk success points to unexploited opportunities in voice, especially in 3G, with differentiated voice quality levels.

  • QOS (quality of service) and measured rates are moneymakers. Odlyzko says encouraging users – and network usage – is the secret to a successful service. Charging flat or simple rates does this, while trying to charge extra for QOS guarantees does the opposite. It adds complexity to the network and confuses users.
Odlyzko also names and shames “a long litany of duds” among recent networking research initiatives. These include: ATM, RSVP, smart networks, active networks, multicasting, streaming real-time multimedia, and 3G. He points out that his pet hate – QOS – is included in all of them.

His keynotes promise an alternative take on a conference agenda focused on helping carriers tackle the challenges they face in moving towards offering a multiplicity of services over a converged, packet-based backbone. The agenda includes sessions on packet voice infrastructure, session border controllers, OAM&P issues, and how carriers can exercise greater control over their converged infrastructures.

To register for one of the NGS 2004 conferences, or to get more information, please click on one of these links:

— Peter Heywood, Founding Editor, Light Reading

Page 1 / 9   >   >>
dljvjbsl 12/5/2012 | 1:17:59 AM
re: Broadband Myths Debunked The article demosntrates that people will pay to have services that will let them communicate. The killer application of the Internet is 'community.' Email, IM, P2P etc. are all manifestations of community.

Note that this belies much of what critics like Postman and Roszak say is the dehumanizing properties of the Internet. Note that it alos belies much of what is said about the need for big pipes to the home.
paolo.franzoi 12/5/2012 | 1:17:59 AM
re: Broadband Myths Debunked
Glad to see this article. Saved me from having to go to the conference now that I know what will be said.

seven
arch_1 12/5/2012 | 1:17:56 AM
re: Broadband Myths Debunked Andrew O. Really gets it.

Users want connectivity. That's it. Not "content". Not real-time. Not QoS.Not multicast. Connectivity, measured in bps.

The history of telecoms over the last 40 years has been user versus provider. The providers desperately attempt to control the users, and the users (abetted by innovative third parties) attempt to work around the artificial barriers imposed by the carriers. This goes back at least to CarterFone.

Features other than bps are merely a desperate attempt at "value pricing." There will always be a way to work around "value pricing" by providing the same value at a lower price, if the underlying costs are low, and there will always be entrepreneurs to help the customer exploit that better way.

Think, people! If you are building services based on Joe Sixpack being too dumb to see through your "value pricing." you are doomed. Would you personally fall for such a trick? No, because you are smart. You would employ a clever work-around. But if you can do this personally, you (or an entrepreneur) can create a company to do the same for Joe Sixpack. (Incidentaly, Joe is not a dumb as you think he is, either.)

Every one of Andrew's "Myths" is based on this holy grail of value pricing. As Andrew points out QoS is a real touchstone. Whenever someone tries to sell QoS, there is snake oikl in the immediate vicinity. (Caveat: where bandwidth is truly limited in a big way, as on a satellite link or where there is an extreme power constraint, QoS might be an issue.)
dljvjbsl 12/5/2012 | 1:17:56 AM
re: Broadband Myths Debunked
The lack of control of network providers over the benefits gained by their customers also points out the futility of the idea of a Class 5 replacement.

The Class 5 has two purpsoes. Firstly it provides services to the dumb stimulus sets (telephones) at the end of the loops. This is its role as a set handler. Secondly, it acts as a traffic concentrator merging traffic from large numbers of cheap telephone loops to a much smaller number of trunks (interoffice connections).

Te development of the smart functional IP set ends the role of the Class 5 as a set handler. The development of the router and IP connectivity ends the role of the Class 5 as a traffic concentrator.

Sets will be served by proxies anywhere in the network. I know personally of an enterprise that is serving its European locations with proxy servers located in North America. The telephone companies' Class 5s and the AIN model associated with them offer no benefit.
rjmcmahon 12/5/2012 | 1:17:55 AM
re: Broadband Myths Debunked Users want connectivity. That's it. Not "content". Not real-time. Not QoS.Not multicast. Connectivity, measured in bps.

More on connectivity, by Bob Frankston, can be found at

http://www.satn.org/about/sepa...

PS. Connectivity is probably measured in something other than bps. Taken from above:

Connectivity - The pragmatic definition: Connectivity is the unbiased transport of packets between two end points.

It sounds like "common carriage" should apply to me.
signmeup 12/5/2012 | 1:17:54 AM
re: Broadband Myths Debunked Tony Li wrote:
"I'd propose something like the sum of (each site * its value to the user * the bandwidth to the site)."

Value to the user is completely subjective and as such cannot be used in any meaningful way within such equation. Numerous attempts in the past have tried to pin down "value to the end user" and failed, taking billions of VC dollars with them.

signmeup
Tony Li 12/5/2012 | 1:17:54 AM
re: Broadband Myths Debunked PS. Connectivity is probably measured in something other than bps. Taken from above:

Connectivity - The pragmatic definition: Connectivity is the unbiased transport of packets between two end points.


I'd propose something like the sum of (each site * its value to the user * the bandwidth to the site).

Tony
rjmcmahon 12/5/2012 | 1:17:54 AM
re: Broadband Myths Debunked The killer application of the Internet is 'community.'

What do you define as community?

How should it be measured? Who should provide its basic infrastructure? Does a society really need it? Who pays for it?
douggreen 12/5/2012 | 1:17:53 AM
re: Broadband Myths Debunked This article brings out a basic problem in the economics of broadband Internet, especially things like fiber to the home. The value to the user is in the content and applications. The cost of bringing bandwidth to deliver that content is in the infrastructure, which is provided and paid for (in the majority of cases) by someone other than the content provider. The value is not coupled with the cost.

Let's take an analogy in the cable TV market. How much would the cable TV providers invest in their infrastructure if they only could charge the user for the transporting of the signal to the home ($10-15 a month), and someone else made money off of the content? Digital TV (and a lot of other things) would not have happened. Their network investment resulted in new revenue opportunities to THEM, not just someone else.

The local carriers are struggling with that equation. They have been unsuccesful in getting a piece of the "content" pie. To try to recover the cost of the network, they created "value added" services which really add little value.

If we don't figure out some way to tie user-percieved value to cost, it is hard to see how we will get an investment in higher bandwidth to the home. Put simply, the carriers have no incentive to invest so that Microsoft can make more money.
HeavyDuty 12/5/2012 | 1:17:52 AM
re: Broadband Myths Debunked "Connectivity - The pragmatic definition: Connectivity is the unbiased transport of packets between two end points."

Connectivity is the physical media (OSI Layer 1)between two points that passes either an analog signal or a data bit stream.
Page 1 / 9   >   >>
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE