Telecom as IT Driver: The Unfinished Tasks

With 100 million telephones being added each year, telecom market in India

has grown at a stupendous rate. India is now the largest telecom market. With

just 10 million telephone lines growing at a rate of less than a million lines

per year, nobody could have even imagined this 15 years back. This major

breakthrough is a result of both: liberalization and technology.

Wireless was envisioned to pave a way for telecom and replace the usual

copper with electronics by several tech experts. Though wireless technology was

very expensive then, but there were speculations that going forward the Moore's

law and the reuse of software will significantly bring down the cost of

wireless. In the year 1994, when the TeNeT group at IIT Madras offered to design

and develop 'Wireless in Local Loop System' within a budget of just 10k, there

were very few takers. But some in the telecom department decided to entertain

this group and provided the initial support. Leveraging the support, the system

was built, commercialized and the belief in wireless increased manifold.

But that's an old story. Now, with telecom being one of the most profitable

businesses in India, and corporates vying to get additional licenses and

spectrum, certain tasks are still unfinished in this arena. But, what are the

unfinished tasks?

Ashok Jhunjhunwala Professor, IIT Madras, ashok@tenet.res.in

The mobile telecom growth in India has been so far largely confined to urban

areas. With urban markets progressing towards saturation, more and more telecom

operators are innovating to break even at Average Revenue per User (ARPU) of Rs

150 per month and deploying base stations to cover rural areas. Most of the

villages of India would see mobiles within the next three years. The real

challenge is broadband. Despite of all its attempts to rapidly increase the

number of broadband connections, the number still stands at a mere 3 million.

The problem is three fold: connectivity, access devices and content (application

/ services). These three areas drive research activities in telecom industry.

Broadband



DSL on copper is proved to be the most cost-effective way of providing broadband
to homes. Unfortunately only incumbent state-owned operators (BSNL and MTNL)

have copper lines for this. The private operators did try to lay copper, but

owing to high cost of deployment and maintenance they failed. The coaxial cable

wires carrying television signals reaches 100 million homes, but is not suitable

for two-way Internet. The only hope for competition is wireless. It has made the

mobile telephony happen in India. But there are a few questions. Can Wireless

similarly drive the broadband? Is wireless technology capable of providing the

bit-rate required for broadband?

Recently, a study carried out by the Centre for Excellence in Wireless

Technology (CeWiT)1 along with Indian operators, showed that India requires

wireless technology to serve at least 700 broadband customers in each cell site

to be profitably deployed. It also calculated (figure on the next page) that to

provide a sustained data rate of 500 Kbps to a user, it would need 30 Mbps and

for 2 Mbps per user, it would need over 110 Mbps in each cell. As operators in a

multi-operator environment, which exists in India are unlikely to get a spectrum

over 10 MHz, the technology required would have to deliver 3 bits/Hz/cell to

over 10 bits/Hz /cell. Today's latest wireless systems (WiMax, EVDO and HSPA)

struggle to provide 2 bits/cell/Hz. They would not compete even with low end

DSL.

TeNeT group is working on wireless technologies to drive this upwards. The

key is to use spatial multiplexing so that multiple conversations are possible

at the same time using same spectrum without spreading. The work combines

multiple-input multiple-output (MIMO) systems, interference mitigation

techniques in multi-cellular systems and robust low bit-rate feedback schemes

for wireless systems to increase capacity. One of the early benefits of the work

would be towards the broadband wireless deployment in India over the next few

years.

Access devices



The second challenge comes from the access devices in Indian homes. Currently, a
personal computer is the sole device that is used. Studies conducted in middle

and lower middle class of the country reveals the fact that though PCs appeal to

these classes, but having a PC in their homes is a daunting task. Cost is one of

the major considerations and the complexity of maintaining is the other. Each

time anything goes wrong, a significant percent of monthly income of these

classes is spent on repairing. Even a simple virus could be a serious problem.

Moreover, they have to get and manage Internet connections too. They need

something simpler.

There can be several solutions to this problem-the one that TeNeT Group has

ideated is a multimedia Network PC, coupled with a managed service. The home

device has no storage and therefore an on/off device. All the storage is

maintained at the server end and is managed by technically sound staff. This

concept is being now commercialized by Novatium, a company incubated by the

group itself.

One may ask how it is different from thin client that has existed for years.

First, the partitioning of the task between the front-end (home device) and the

back-end server is the key. One would like to do as much as possible at the

front-end, thereby considerably reducing the load at the back-end (as the

back-end is shared). Secondly, for the concept to truly scale, the front-end

would not be any special hardware. A simple set-top box or a cellphone with some

special interfaces or downloadable software will be adequate. As these devices

are usually made of a general purpose Digital Signal Processors, all the

multimedia and graphics would be handled by the front-end. For a normal home

user these tasks consume most of the CPU resources. Bringing them to the

front-end makes the servers handle more number of users. Making this

partitioning totally independent of the user program development so that any

software at the server automatically partitions itself is another big challenge.

Content: Rural focus and services



The broadband connectivity discussed above can enable broadband connection in
6,37,000 Indian villages, where 700 million Indians live. The network PCs would

enable multiple terminals at Rural Business centers. However, the key remains to

use IT services to benefit villagers, and the priorities are education, health

care & livelihood.

Today in rural India, there are few good teachers. Educating a large number

of children living in rural areas in such a situation is a challenge. It is

possible to use IT to partially compensate for this shortcoming and supplement

the education for the children.

IT and Communications are powerful tools that society can use to transform

itself. However, they are merely tools and not an end. Research in IT needs to

recognize this and should continue to use Telecom and IT to strengthen Indian

society.

1 CeWiT is a public-private organization driven by the TeNeT Group and

currently located at IIT Madras.