
•
MANOU MARZBAN
ON ADSL - THE PATHWAY TO BROADBAND?
Basically, ADSL makes your Internet experience faster…and that
is something we all want.
What is all this
talk about ADSL in the UK?
You probably have read or heard something about ADSL by now. It’s
the sort of thing that makes grown people with a background in technology,
giggle excitedly like school kids. Advocates of ADSL claim it will change
how we use the Internet and therefore, they claim, it will change how
we communicate in general. My perception is a bit more cautious, only
because technology moves so quickly that before we can even understand
the basics of ADSL, a new acronym will be introduced that promises to
be better, faster and cheaper than ADSL. Nonetheless, you need to know
about ADSL, as it is about to be plastered over everything Internet
in the UK.
For those new to ADSL, the acronym stands for Asymmetric Digital Subscriber
Line. What ADSL basically does is to transform existing twisted copper
wires between a local telephone exchange and a customer's telephone
socket into a high speed digital line. It is called "asymmetric"
because it moves data more quickly from exchange to customer than from
customer to exchange. This makes it particularly suitable for applications
where customers expect to receive more data than they transmit such
as use of the World Wide Web, corporate intranets, and reception of
digital audio-visual material.
How does ADSL work?
Although ADSL operates over the same twisted pair copper loop used to
provide voice telephony, it utilises the frequency band from 20kHz to
1.1MHz. The scheme, first implemented by AT&T, is called Carrierless
Amplitude Phase Modulation (CAP). However, before AT&T could deliver
significant quantities of chipsets to roll out ADSL in North America,
the American National Standards Institution (ANSI) created a standard
(T1.413) for ADSL based on a technique known as Discrete Multitone (DMT).
Are you bored to tears yet? Wait…it gets better.
DMT operates by dividing the available bandwidth into 32kHz wide "bins"
each capable of carrying upto 64kbits of information. This allows a
theoretical maximum transmission rate of approximately 16Mbps, but current
chipsets have a ceiling of 12Mbps. This is split into two parts, "upstream"
to the telephone exchange and "downstream" from the exchange.
Because of the way bins are allocated, the maximum upstream data rate
is approximately 640kbps, while downstream performance is approximately
8Mbps. Hence the technology is suitable for applications such as the
Internet, where you expect to receive more information than transmit.
As I mentioned, ADSL operates over a traditional telephone line. This
means that an existing telephone line can be used to send and receive
high-speed digital data. The ADSL signal is carried by two ADSL modems
- one at the user end and one in the local exchange. These ADSL modems
are designed to exploit the physical transmission capabilities of the
copper line, to achieve higher data rates over the line than ever before.
A 'splitter' (which is a filter), one at the user end and one at the
exchange end, separates the telephony signal from the ADSL signal. This
means that telephone calls can be made at the same time that data is
being sent or received (i.e. a customer can surf the Internet and still
make telephone calls). This is a true bonus - for those of you with
teenage children - you know what I mean. The benefits of ADSL are simple:
1) Speed - its high speed data rates provide great potential for both
commercial and domestic use, making video-on-demand a reality, 2) Always
On - customers no longer have to waste valuable time obtaining a connection
via dial up, and 3) Flexibility - it does not affect the normal telephone,
so customers can make and receive telephone calls or faxes whilst they
are on line.
So, with all the technical jibberish out of the way, what does this
mean to the punter? Basically, ADSL makes your Internet experience faster…and
that is something we all want. Depending on line length between the
exchange and the customer, it is possible to transfer data downstream
at up to 6Mbps. This is 10 to 40 times faster than today's typical modems.
What this all means is that it may be possible that ISDN, lately offered
as an Internet and remote access solution, may be rejected in favour
of ADSL. There are some issues. For example, data transmission performance
reduces with increases in the distance over which the information is
transmitted. In short, this means that only subscribers close to the
telephone exchange (central office) will receive data at rates sufficient
to carry MPEG-1 or 2 video, but almost every subscriber will be able
to obtain data transmission rates of 1.536Mbps, which I'm sure you will
agree affords excellent internet performance. Certainly better than
what we currently have on the books anyway.
What is going to
happen with ADSL?
As the years, or, more likely months pass, there will be a multitude
of services striving to use bandwidth on the Internet. We're all familiar
with analogue modems. They've been around in one flavour or another
for many years, for example, dial-up or leased lines. Everyone also
knows that even 56Kbps (even if your phone line and ISP can actually
achieve it) is not going to be enough bandwidth to fully take advantage
of the Internet. Many new sites contain multimedia performances (Java,
Flash) which are just too painful to wait for if you have to download
them at 56kbps, or even worse....33.6kbps! So what are our options as
consumers?
Urge your network operator to install optical fibre to your home as
soon as possible. This may be to no avail…as the most probable
answer you will receive is that fibre is too expensive to deploy on
a countrywide basis. Once you have been told this - you can say 'fair
enough - how about ADSL?'
Network operators have too much investment tied up in the copper loops
which already join their exchanges to our homes and offices to give
it up easily. However, since ADSL technology incorporates the existing
copper infrastructure - network operators will be keen to adapt to ADSL
because it extends the life of their investment by many years. Their
problem is that if they give us a quantum leap in performance now, it
will be a very short time before we ask for the next step. However,
increasing use of the Internet for all manner of tasks mean that we
should demand ADSL now, because, with the best will in the world (take
ISDN as an example) deployment will take at least two years. So there
is no harm in contacting your local operators and asking about ADSL
connectivity. What will it cost? Who knows…initially, I would say
in the £300-£400 per year range - as in current ISDN rates.
Call BT and see what they charge, always a decent benchmark.
There are hurdles. No authority has, as yet, published a document which
describes the filtering arrangement necessary to separate the voice
band signals from the ADSL band frequencies. The filtering arrangement,
known as a POTS splitter will need to be addressed shortly if a delay
in the commercial roll-out of ADSL in Europe is to be avoided. The reasons
for this are quite simple.
Firstly, in the US and Canada, a 600 ohm termination of the telephone
line is the norm. This permits the use of a very simple, passive splitter.
Because of the complex termination arrangements in Europe, such a splitter
cannot be implemented in splitter designs for ADSL in Europe. Secondly,
the European standards for telephone line use do not anticipate use
of the frequency bandwidth above 100kHz. Add the complication of inserting
another device into the telephone network in series with the customers
branch system and the situation becomes even less clear. The ADSL Forum
is taking steps to address this issue, but no hard and fast measures
have so far been taken to remove these obstacles of confusion which
must confound the system engineers in companies such as BT and Kingston
Communications.
Finally, until the standardisation issues are solved - we will have
to make do with ISDN or worse. But watch this space, for it is only
a matter of time before ADSL becomes a reality here - as in the United
States. Once that happens, your entire perception of the Internet, surfing,
ecommerce and Web communications in general will change. That is a promise…
for we all know one thing…time is money and speed is everything.
Manou Marzban is Head of Content and Channels at BiblioTech, 50 Carnwath
Road, London SW6 3EG, 0171 384 6900, he can be emailed
at manou@bibliotech.net.