
Why your broadband's not as fast as it could be
Published: 8 June 2004 07:40 BST
We have the technology to provide greater bandwidth than is currently available. But that doesn't mean providers are courageous enough to offer it, says Martin Brampton.
Cisco has launched a new internet router claimed by the company's chief executive to be "the biggest jump in innovation since our first router 20 years ago". It is said to offer a hundred-fold increase in capacity and can support 40Gbps internet connectivity. What can we do with this speed?
To put it into context, a good-sized ISP might have 10 internet connections at 155Mbps each; a small one might have only one. Or for another comparison, Linx, the London Internet Exchange that carries a high proportion of UK traffic, hit a peak of 25Gbps last Christmas.
These are awe-inspiring numbers and one wonders what impact they will have on the communications world. On the face of it, technology developments stand to reveal again the disparity between what is achievable, and what, for commercial reasons, the major telecoms providers choose to make available.
As with so many other aspects of IT, the dot-com boom created enormous distortions. The telecoms operators and their major suppliers achieved share prices that suggested that simply carrying data was a major part of the economy. The subsequent slump revealed the obvious truth that the carriage of data is an important service, but not in itself a major economic driver.
Operators have generally followed BT in a rearguard action to keep prices high. Indeed, most providers do not have the luxury of choice, since many services rely heavily on wholesale services provided by BT. For example, the vast majority of broadband services sold through a host of independent providers are based on the standard BT wholesale offering.
Dramatic increases in bandwidth are available. We all suffered as telecoms companies dug up roads to install more and more fibre. Much of that is lying idle. With large improvements in the speed of the optical devices used to drive fibre optic cables, each cable has been able to carry much more traffic than originally anticipated. As a result, many installed fibres do not even have terminating equipment.
Now with the introduction of a huge hike in the capability of routers, there is no technical obstacle to radical changes in the availability of bandwidth. A major barrier, though, is the reliance of operators on voice traffic for revenue. Inevitably, everybody is now aware that voice traffic is essentially just another kind of data. Current charges for voice calls are illogical and unsustainable in an environment of rapidly falling costs for data transmission.
This foot-dragging is likely to have a damaging impact on broadband. The great technical achievement of broadband was to raise data rates over the 'last mile' so that the vast majority of telephone subscribers could access high-speed services. For the time being, users are finding that broadband gives a dramatic speed boost over dial-up.
This will not last, though. As the number of users increases and at least some of them start to make heavy use of high-speed data transmission, the contention factor will kick in. Typical domestic broadband has 50 users contending for 2Mbps, or an average of only 40Kbps per subscriber. The whole reason for this is throttling of the service when it reaches the telephone exchange. BT naturally uses the contention problem as a sales tool for supporting its leased-line products. Yet the tools to hugely increase the capacity of the links through and between exchanges are available.
We therefore remain in the traditional situation, where local data communications are far faster and cheaper than wide area communications. This is a pity as it is a severe constraint on IT developments. It is my belief that cutting the cost of data transmission would result in an increase in consumption that would more than offset the price reduction. In other words, more money would be spent. Unfortunately, unlike the suppliers of most IT equipment, the providers do not have the courage to take this path.
Martin Brampton is founder of Black Sheep Research, an independent consultancy providing research, writing and speaking services on a wide range of business and technology issues. Martin was previously a director at Bloor Research, and has worked with IT as a user and analyst for over 20 years. He is a longtime contributor to silicon.com and his blog can be found on his website.
Take a lead role in problem determination and resolution when they do occur, with the provision of (paid) out-of-hours support and disaster recovery ...
You will be working in one of the teams that engender these high levels of trust by delivering: Innovative technical developments High speed ...
Want to join a company with a proud and long heritage in Northern England where you will be the manager of the SAP Finance section with ...
CIO50 2008
The silicon.com CIO50 2008 profiles the most influential and innovative tech chiefs in the UK across all industries and organisation size, from the biggest FTSE100 companies to high growth dot-com start ups and the public sector. The list was voted on by the UK CIO community and a panel of experts. Find out more in our latest special report.
Stories from the web...
Copyright ©1995-2008 CNET Networks, Inc. All rights reserved. Top of page
Natasha Lomas iPhone nano: Fact or fiction? Apple's game will keep us guessing...
Peter Cochrane Peter Cochrane's Blog: New niche for 3G? ADSL costs could drive 3G uptake...