'Some Network Functionalities Will Move From Electronic To Optical Domain' - Atul Srivastava of Bell Labs Photonic Networks Research department

alt="Atul Srivastava" align="right" hspace="0" vspace="3"

width="118" height="149"> size="2">... bps, kbps, Mbps, and now Tbps. Scientists from Bell

Labs (where else?) have demonstrated the world’s first

long-distance transmission of a terabit of information per second

over a single strand of optical fibre. Which means a system based

on this technology can carry the whole world’s Internet

traffic over a single optical fibre, simultaneously.

Using an experimental ultra-wideband

optical-fibre amplifier, unveiled less than a year ago, the

scientists were able to transmit signals at the rate of 10

gigabits (billion bits) per second over each of 100 wavelengths,

or colours, of light for 400 km, or nearly 250 miles. Though in

the past two years, researchers at Bell Labs and other research

labs have demonstrated terabit-capacity transmission over

relatively short distances, this was for the first time that a

long-distance transmission was demonstrated.

A research team, led by Yan Sun and Atul

Srivastava, of Bell Labs Photonic Networks Research Department,

presented a technical paper on the 100-channel long-distance

terabit experiment in the Optical Fibre Communications (OFC)

conference in San Jose.

In an E-mail interview with Shyamanuja Das,

Atul Srivastava tells how they did it, and makes some

prophesies about the future.

This is a great achievement.

Did you use some new material or new fibre design?

The terabit transmission

experiment was made possible by the invention and development of

80 mm bandwidth optical amplifier in our laboratory. The

amplifier has enough bandwidth and output power to support more

than 100 optical channels. In the experiment, 100 optical

channels each carrying 10 Gbps data were used through a chain of

four fibre spans (90-110 km each) and four optical amplifiers. In

this demonstration, no new material or new fibre design was used.

Is there any roadblock in

turning the "experimental" breakthrough into a

"commercial" triumph? I mean technological,

environmental, market ... any block? If yes, what?

We do not see any technological

roadblock in turning this experimental demonstration to

commercial system. In fact, the parameters for the experiment

(e.g. 10 Gbps bit rate which is standard SONET/SDH OC-192 rate,

100 GHz channel spacing which is ITU-standard, 2^31-1 PRBS word

length and 80-100 km fibre spans) were chosen to be close to that

of current terrestrial communication systems.

So far, you have played on

bandwidth/speed. Now, despite faster technological growth in

fibre, you have to go to electronics to do certain jobs. That

might be a constraining factor ...

In this experiment, we

increased transmission capacity limit by enhancing the optical

amplifier bandwidth. We believe that in future some of the

network functionality will be transferred from electronic to

optical domain.

What is next from Lucent?

Lucent has already announced

400G wavestar system, which will be available by the end of this

year. This system will have 400 Gbps capacity with up to 80

optical channels. Considering the demand for higher capacity, we

think that higher capacity systems are inevitable as we go

forward.

When will this technology

come to, say, campus-networks and LANs? And what are the

challenges?

WDM technology is already

making inroads into LANs, WANs, and MANs. We will see larger and

larger numbers of WDM systems in these areas as well. The biggest

challenge in these systems is to keep the cost very low. We

believe that the ultra-wideband amplifier demonstrated in this

experiment will be the enabling component of the future optical

networks.