Real Broadband Access
by Stephen Lowe

In whichever consumer Internet magazine you care to open you will find an article on broadband. It will, most likely, complain of the lack of availability and will denigrate the telecommunications service providers for their lack of commitment to always on, virtually free high speed Internet access over which users may share pirated versions of audio and video products. So, is there any substance to these complaints and if so how could they be resolved?

The worlds of entertainment and of communications are, it is said converging.  There are those who believe it will be a quiet evolution. Others may share my view that it is a confrontation.  Like stars colliding the two worlds of broadcasting and Internet will meet and from the consequent explosion will come a new industry - digital interactive services.

A traditional view of the broadcast and Internet value chains shows content creation at one end and a passive viewer at the other.  The products are delivered down two separate networks, one, the broadcast chain, with high capacity in the downstream direction but no up stream channel and the other, the Internet, with the addition of an up stream channel but with limited down stream bandwidth.  At the end is a viewer who merely consumes the product as it arrives.   Control in this model resides with the content and service providers.  The consumer may feel they have a choice available but in practice it is limited to what the market analysts of the providers think can be sold with a reasonable margin.

With the increasing availability of digital content, music, video, graphics and games, the power has begun to move away from the traditional models.  Anything a television broadcasting channel wants to do can be done over an Internet service provided that the access is fast enough.  Once there is a high speed connection in place then voice telephony is cheap and simple.  And, most dangerous to the traditional media companies, music and video can be duplicated with no loss of quality at will.  And, by sending it over a high speed Internet connection, it can be shared on a world wide scale for virtually no cost.

In this world the consumer has taken control and is deciding what and when to consume.  They have moved from viewer to a combined Viewer and User, A "viewser" as someone coined the phrase.

To make this happen in a consumer market requires two things.  First, a connection that, like a broadcast, is always on. Second, there must be adequate bandwidth for the consumer to get the product instantly when they want it.

This then leads to two conclusions.  There will be a common transport protocol for all types of content and the current telecommunications and cable networks will have to increase the bandwidth available but at the same or a reduced cost.

How can we do this?

Satellite is the traditional home of multi-channel linear television.  Linear television is of course just another name for broadcasting where one programme follows another is a pre-defined sequence.  We are seeing the migration towards interactivity, or non-linear broadcasting, and we have a new phrase in the lexicon - "press the red button".   What we get when we do is access to other channels and the feeling that we are in an interactive world.  But without a return path what we are actually using is additional channels that are not otherwise accessible.  In order to support the email and other true interactive services that satellite promises there has to be a partner with the return path and this is of course a teleco with a standard dial up modem.

The most widely deployed alternative is the family of Digital Subscriber Line technologies beginning with Asymmetrical DSL.  By using existing telephone copper access networks ADSL allows the service provider to roll out rapidly and widely with network upgrade work limited to the nodes, or exchanges.  A business plan that balances the cost of the exchange upgrade against the aggregate subscription revenue in the service area is relatively easy to work out.  At present subscription rates it requires between one hundred and one hundred and fifty high use subscribers, perhaps SMEs, to offset the capital costs. 

Once ADSL has addressed the initial market then we will see the growth of higher speeds and a move to symmetrical services.   To achieve this will require a good deal more network engineering upgrade investment because the range of VDSL is significantly less than ADSL.  Current proposals include the installation of street hardened equipment, perhaps in below ground housings, generally within a kilometre of the end user.

The other widely deployed network is of course the cable networks.  While penetration of cable varies very widely between countries where it is available it has the potential to deliver more than either of the previous options.  Cable networks rely of a deep penetration of fibre, and hence high capacity, close to the end users.  In a typical cable access network it is possible to contemplate an offer of 30 Mbps downstream to every home on the network.  The upstream bandwidth is less easy to provide because of the nature of the frequency allocation plans on the fibre and coax.  However, it will still allow a far higher speed than for the current implementations of ADSL.

Until this last year it was assumed that the shining knight riding to the rescue of the population starved of broadband would be wireless.  Broadband Fixed Wireless Access spectrum has been auctioned across Europe and of course the Americas.  It has been successful in very specific ways.  By providing access where there was no infrastructure in emerging economies and in high density urban populations where by taking a small percentage of the market an operator can make a profitable business.   The most typical frequencies in use are the 3.5 GHz, 26 - 28 GHz and, in some areas the 38 GHz bands.  Other bands, 40 GHz and higher, are in development but suffer from high hardware costs for the consumer equipment.

At even higher frequencies we are seeing a steady increase in the deployment of Free Space Optical systems.  FSO offers a number of advantages, no spectrum licence fees for example, but is limited in range to between 500 and 1,000 metres.  Fog and mist are the most damaging weather conditions and a rule of thumb that seems to be emerging is that a FSO system will carry about one and a half times as far as you can see.  The other limitation to FSO at present is that it is a point to point only technology. However, it is probable that this will change in the next year or two.

Where the greatest excitement can be seen this year is in the growth of the licence exempt wireless equipment, commonly know as WLAN or Wireless Local Area Networks.  The two main bands in use are 2.4 GHz and the 5.2 GHz although these are sub-divided into bands for specific uses.  The term licence exempt refers to the condition that allows anyone to operate a WLAN provided that they use type approved equipment.  Volume production has brought the products within the scope of domestic users and we are seeing massive growth in the deployment of community wireless networks for both academic and social groups.

The attraction of licence exempt technology is that the most costly element of the network is moved from the service provider's books to the consumer.   Much in the same way that we no longer expect our telephone company to provide the actual telephone; the subscriber is encouraged to buy the instrument to put on the end of the access network. This has led to the creation of a new industry and greater consumer choice.

All that the service provider need do it set up a suitable wireless hub and grant access to anyone with a suitable wireless equipped laptop or PDA.  The race is on to capture the business user market by providing access points in railways stations, airport lounges, hotels and high street coffee shops.  Such networks may be supported by an ADSL, BFWA or cable modem broadband connection to the trunk.

Regulation in some countries does not yet permit commercial public access using the licence exempt bands but this will have changed by the end of the year opening up a global opportunity to both hardware vendors and service providers.  It is also clear that the transport will be based on the Internet Protocol since this is already a business and consumer standard.

At risk from this technology is the growth of the 3G mobile services.   Since a high percentage of the service functionality that a business user requires may be provided by the low cost WLAN technology there has to be a clear driver to support the business plan for high hardware cost 3G. One possible driver may be that the unrestricted use of licence exempt systems will result in such spectrum congestion that systems cannot provide the Quality of Service that a business user requires.  It may therefore be that the licence exempt systems is a transitional phase that develops the market for mobile broadband ahead of the roll out of 3G.

On the horizon too is the Ultra Wide Band technology that has been developed over the last couple of years and which was given a first licence in the USA in Spring this year.  The European market is looking to see how the FCC licence process develops before setting out a standard so we may expect there to be two or three more years to go before UWB enters the market place.

In summary, broadband growth depends on three elements coming together.  There must be a common transport protocol for all types of content and this will be IP.  The network or service provider must drive the cost of the consumer unit off their books and let the consumer pay. New partnerships will have to be formed to match the new value chains that get content from creation to consumption.