Bandwidth has been a widely discussed topic in India in
recent times. On its availability hinges the country’s ability to realize the
dream of $50 billion of software exports, a $87 billion software industry and a
$140 billion IT industry by 2008. Today, the total bandwidth available to the
country stands at 320Mbps, routed through VSNL. This capacity is expected to
double by September. However, the country’s need, according to a Nasscom
report, stands at 10Gbps by the end of 2000, and is expected to go up to 40Gbps
by 2001. In view of this, Nasscom has put forward the notion that every
individual has a right to demand 2Mbps of bandwidth as a basic need.
The government has initiated steps to ease regulations and
facilitate the availability of bandwidth in the country. As many as 65 gateway
licenses have been issued through which submarine bandwidth providers can now
sell their capacities to users directly. Even then, the fact remains that India
will continue to face a scarcity of bandwidth till 2002. Says Sharat Jain,
country manager, India, Teleglobe, "Irrespective of the new capacity being
made available, India will be short of capacity by a factor of four till the end
of 2001. Things will improve by 2002 once new cable systems are in place."
Adds Harpreet Duggal, VP, sales and marketing, BT-Worldwide, "There are two
aspects to the bandwidth issue. One is the international connectivity, which
will not be available to us until the landing stations by submarine cable
operators come up. That can easily take up to 18 months. The other aspect is the
domestic connectivity which again is a scarce commodity."
In fact, it is this domestic scarcity where the bottleneck
lies. Although the national long-distance service has been opened up and players
have joined the race for providing capacity, none of them are geared up for
large-scale commercial deployment of their services in the near future.
In this scenario, if one decides to procure capacity from
submarine bandwidth providers, multiple landing points would be required, which
would be very costly. Besides, even that would be possible only in coastal
areas. Other places will continue to suffer because the capacity to those places
would be routed through domestic cables from the landing points.
The orbital alternative
Satellite gateways can be set up anywhere. "One
alternative solution is to allow the interconnection of VSATs to other
networks," says Rothin Bhattarcharya, managing director, Telecompetence
India. Installing satellite gateways close to an area of demand and then
distributing the capacity over the terrestrial cable can be a reasonable option
as it would be cost-effective and also help create a public network
infrastructure. In the long run, perhaps things would move to terrestrial
networks anyway. At present, however, such an approach would favor only big
corporates who can advantageously afford the expensive VSAT networks.
For ISPs, two factors favor the choice of satellite-based
international gateways in the short run. One, the time to set up, and
consequently the time to market. Two, the relatively low cost of the initial
set-up. Says Bhattacharya, "The roll-out of satellite-based connectivity is
quick and therefore a definite plus from the time-to-market aspect."
Moreover, below a threshold point, there is no difference
between the cost of the satellite bandwidth and that of the submarine bandwidth.
Says Jain, "It is only when huge volumes are purchased that the submarine
bandwidth cost becomes cheaper."
Satellites are increasingly being advocated as a clean
solution in the short run. Later, when general-purpose users migrate to the
submarine option, the role of satellites would become more specialized. Says
Amitabh Singal, president ISPAI, "Ultimately, satellites will be used more
as a back-up medium when more and more mission-critical applications go
online."
Cables are certainly superior in the latency aspect–the
time required to connect between two stations. Cables have an average delay of
100 milliseconds while in case of satellites it is 500 milliseconds. This aspect
of latency is especially significant when the transaction protocol is IP. So, as
we move to IP, submarine cables are likely to become imperative. But till then
satellites appear to be the most preferred choice.
Balaka Baruah Aggarwal
in New Delhi