WIRELESS INTERNET:Cable Run

Broadband in India has entered a definitive phase. From times
of 'to be or not to be', we are entering a phase of aggressive marketing by
most broadband companies. The market is hot. TRAI fired the first salvo and
defined the term broadband. Operators like BSNL picked up from there and brought
the prices of broadband connectivity crashing down. This in turn has forced
competitive carriers like VSNL and Bharti to push down the prices. But despite
all this, neither has the net usage increased dramatically nor has the
competition been able to bring in any large numbers of broadband customers. If
anything, the ARPU for operators have decreased.

In this background, key questions are, why do we need to
promote broadband and how do we promote it?

Broadband and GDP

Broadband connectivity has been positively linked to a nation's GDP-better
the broadband connectivity, better the GDP. However, can an always-on, 256 kbps
connectivity by itself raise the GDP?

Broadband connectivity can be really useful if it can
decrease the distances in real terms and eliminate the cost of travel. It has to
make available services like e-health, e-agriculture, e-shopping, e-gaming, and
e-governance by making services connect to people easily and in more affordable
ways; and provide newer and better services to enable higher productivity. Only
then would broadband contribute to raise the GDP.

Towards this end, broadband has to provide video services
apart from voice and Internet. More, it has to provide interactive video
services to ensure the all the enabling applications, mentioned above, become
possible. It will also mean that broadband would have to provide bandwidths much
higher than the 256 kbps, because many applications are gong to require much
higher bandwidths. And the bandwidth will have to be dedicated for each user
because shared-bandwidth services cannot provide such bandwidths.

50 million broadband and Triple-play connections

As a nation, we should have a target of reaching at least 50 million
broadband connections. In a population of 1 bn, 50 mn is just 5% of the
population. In terms of households, if we assume on an average five persons per
family, it roughly means just about 1% of the households.

Questions remain. Can we really reach 50 mn broadband
connections? What kind of technology can take us past the 50 mn number and
provide real triple-play services that can impact the GDP?

While considering the possible solutions for broadband, like
copper, wireless and cable; in the Indian scenario, the one with a lot of
potential could be the coaxial or the community access TV (CATV) cable. Its
subscribers have telephone subscribers, and it is still growing. This cable has
the bandwidth capabilities to provide interactive, triple-play services.

In a way, today 60 million homes have broadband access of 2
Gb in the form of CATV cables. However, the CATV wires in India are not
bi-directional. If there was a solution that could ride on this network, without
calling for much change in the infrastructure or the operator environment, it
could facilitate very fast broadband rollout.

What Ails CATV in India?

CATV infrastructure in India came up due to the entrepreneurial energies of the
youth. The aim of this cable network was to distribute entertainment and
information channels at the lowest possible cost. Hence all the network
components, including the cable and connectorization happened with the
lowest-cost method. The aim was to make it work, rather than professional
network installation.

And it has not done too bad till now. The neighborhood local
cable operator (LCO) was always available to the customer. Hence, the service
they provide is good. Things can fail, but they will be rectified immediately.

CATV in India has worked without proper terminations, with
low-quality splitters and tap-offs, nonuniform low-cost amplifiers, and little
technical knowledge.

And these LCO's have still ensured ensured that this
network has grown in terms of numbers, stability, quality of service, and
customer satisfaction.

But the fact remains that it a one-way, overhead coaxial
network, which was set up on treetops and buildings, and is vulnerable to the
elements of weather and earthing issues.

Converting this network into a two-way network is a
multi-dimensional challenge. Conventional cable modem termination systems (CMTS)
use sub—50 Mhz on the cable for upstream. However, providing upstream on CATV
in India would require-among other things-network components like amplifiers
to be changed and connectorization at various places being made proper. The SNR
as the noise goes on adding up, as the subscribers increase (in spite of all the
balancing) is a severe pain. And because it is an overhead network, it is
susceptible to noise pickups, which would make the two-way cable network
difficult to sustain and grow.

Another problem is the lack of adequate technical training
among the LCOs. And given ratio of subscribers:LCOs and the costs that need to
be put in, two-way in CATV becomes an unviable entity in India.

Downstream on Cable, Upstream on Wireless

Most of the problems in making overhead CATV cable a two-way network can be
solved, if we provide the upstream on wireless. Even if the upstream bit rates
are not as high as the downstream, it is still possible to provide good quality
triple-play services on the network.

The challenge in this is to segregate the traffic at head-end
and combine it at the subscriber terminal. Fortunately, IP traffic lends itself
to be split into two different paths. Issues like latency and effective
throughput need to be monitored to ensure that scheme of downstream on cable and
upstream on wireless works.

40 Mbps on a Single CATV channel

ITU J.83b standards define the way in which 40 Mbps data can be transmitted on a
single CATV channel of 6 MHz bandwidth-40 Mbps can be got with 256 quadrature
amplitude modulation (QAM). The conventional CMTS uses 64 QAM and provides near
30 Mbps on a single 6 Mhz Channel. Typically they work at signal levels as low
as 45 dBuV.

If the Indian CATV cable network is going to continue as
one-way, there would be no problem in setting up the signal levels much higher
then the 45 dBuV, because problems of reverse path noise would not exist. With a
60 dBuV signal, the CATV network can support 256 QAM and this would provide 40
Mbps on a single channel.

TV Works Internet Works

Another key issue is that the end subscriber as well as the LCO should be
able to identify problems related to running data on cable, easily-if Internet
does not work on cable, there should be methods effective methods for finding if
it can work or not. And this system needs to be user friendly.

Assuming that upstream goes on wireless, for downstream on
cable the parameters that one would be looking at to ensure that Internet on
downstream channel is okay, are: signal level, bit error rate, and modulation
error rate.

A signal level of 60 dBuV for ensuring 256 QAM on CATV
network in India would require amplifiers, that are present in the network and
in most cases are easy to set up. A bit error rate of at least 10-6 is required,
while ideal requirement is 10-8. Bit error rate can be influenced by the signal
level and the interference level. Since signal levels are easy to maintain, the
interference levels are the ones to look at.

Interference can originate from the head-end or amplifiers or
unterminated connections. Of these, ensuring minimal interference at the
head-end is easy. Most LCO's do it daily. Ensuring minimal interference at
amplifier points is bit more tricky, but this is also a job that the LCO does
routinely. Unterminated connections causing reflection and interference is a
fact of life in India, and is bound to cause higher bit error rates. This
problem can be solved either by ensuring that all open ends are terminated
properly or by placing the channel for Internet at a higher frequency. At
frequencies above 300 Mhz, the attenuation on cable is high and hence the
reflections from unterminated connections attenuate rapidly, and are unable to
cause any significant interference.

Hence it is very possible to set up a 256-QAM,
digital-Internet channel on a 6 Mhz CATV channel in India. And we can ensure
that if the TV works, the Internet also works. There could be conditions in
which the TV would work even with slight problems, but by ensuring adequate
signal level, higher frequency, and terminating the open connections we would be
able to ensure that Internet works when TV works and TV also works with good
quality picture.

R Balajee, head

cable wireless business, Midas


vadmail@ybermedia.co.in