OPTICAL COMMUNICATIONS

With the emergence of low-loss glass fibre as a broadband medium for transporting voice, video, and data traffic, fibre-optic technology has proven itself as a crucial milestone in the global telecommunications and IT revolution. As technological advances in photonics and electronics permit more and more wavelengths of light to be transmitted at ever-increasing data rate, the data carrying capacity of each strand of optical fibre is doubling every year. Transmission fibre has today evolved from a simple light pipe to a finely tuned medium that not only guides light but manipulates the optical pulses as they propagate.

A typical loss spectrum for an optical fibre is shown in figure which indicates the minimum light attenuation of about 0.2 dB/km at 1550nm wavelength. This corresponds to loss of about 5 percent of the light after propagation through one kilometer of fibre, implying that only one in a trillion photons leaving US would reach Japan. The optical losses due to absorption and scattering required electronic "repeaters"—placed at every tens of kilometers—to amplify the signal along the fibre to compensate such losses. Each repeater detected signal, filtered it and retransmitted it using a new laser diode.

A major revolution in the industry has been brought about by Erbium Doped Fibre-optic Amplifiers (EDFA), invented in 1987. These amplifiers permitted the direct amplification of optical signals—without conversion to the electronic domain—allowing propagation within a fibre for hundreds of kilometers using periodic optical amplification (every 40 to 80 km) to compensate for the fibre loss. These amplifiers are more reliable and generally less expensive than electronic repeaters, with their performance independent of bit-rates. They support more than one wavelength enabling WDM (Wavelength Division Multiplexing) systems and networks, increasing their capacity many times.

These optical amplifiers are based on an optical fibre doped with a fraction of a percent of the rare-earth element "erbium". These erbium ions absorb light at 980 or 1480 nm (nano meter) and store this energy to amplify signals at wavelengths near 1550 nm. In a wonderful coincidence of nature, this amplification band coincides with the minimum-loss wavelength of optical fibres. Moreover, EDFAs are easily pumped, have low noise, and unlike semiconductor amplifiers, do not generate cross talk between channels.

Conventional band (C-band) EDFAs amplified 1530-1562 nm band, which is also quickly being consumed. The next-generation EDFAs are focussing on Long-wavelength band (L-band) of 1570-1610 nm.

While erbium amplifiers have been tremendously successful, they are only capable of amplifying 80 nm of the low-loss window. The spectral bandwidth of conventional EDFAs is beginning to be a capacity limitation. To fully exploit the capacity of optical fibres, new amplifiers that can amplify signals from 1270 to 1670 nm of wavelengths are needed. The only optical amplifiers that work over this wavelength range are "Raman amplifiers", which are now reviving after years of hibernation, particularly after the developments in the area of high power diode pump lasers and the optical grating technology. "Raman Scattering" being anon-resonant process, can provide amplification for both 1300 nm and 1500 nm windows.

The advantage with these amplifiers is that any optical fibre can serve as the amplifying medium. By simply injecting intense pump light into an existing transmission fibre, signals propagating within the fibre can be amplified. Since Raman gain occurs in all optical fibres, even installed transmission fibres can be made to amplify communication signals by injecting pump light into the fibre span. In essence, the transmission fibre itself becomes the amplifier medium. Launching pump light into the end of the transmission span and allowing it to counter-propagate relative to the signals, triggers distributed Raman amplification, which can greatly enhance the performance of communication systems.

In Raman amplification, an intense pump light couples to vibrational modes of the glass of fibre itself and is re-radiated at a longer wavelength, amplifying a signal if the pump is at an appropriately shorter wavelength than the signal. The gain spectrum is determined simply by the pump wavelength rather than the fixed energy levels of a dopant like erbium as in EDFAs. Since the Raman gain spectrum is not locked to fixed energy levels like erbium, a desired Raman gain can be generated at any wavelength in the infrared,as long as the requisite pump light is available. This feature allows Raman amplification to be applied across the entire transmission window of silica optical fibres.

Raman amplifiers have been demonstrated at 1300, 1400 and1500 nm with gains as large as 40 dB and noise as low as 4.2 dB, and are the only silica fibre amplifiers to demonstrate analog grade noise performance at 1300 nm, which is of great interest for CATV signals. These amplifiers have been tested as high-gain, low-noise optical preamplifiers for Dense Wavelength Division Multiplexing (DWDM) digital transmission systems and have demonstrated 80 Gbps transmission for over 140 km of fibre as well as DWDM systems at 1400nm. These have also been used in combination with EDFAs to show transmission of 40 wavelengths, each carrying 40Gbps over 400 km of fibre.

Ultimately, with the present ongoing developments, it would be possible to generate a laser—with any light in the infrared between1100-1700 nm—and use it to pump a Raman amplifier over the entire transmission window of optical fibres. And that might be just the beginning of full realization of the capabilities of Raman’s gift to telecommunications.

Raman Effect

The phenomenon of "Raman Scattering" was discovered in 1928 by Indian physicist, Chandrasekhar Venkata Raman, popularly known as CVRaman, who received the Nobel Prize for Physics in 1930, and was later awarded the "Bharat Ratna". The first experiments demonstrating Raman Scattering were conducted with extremely limited or, if one can say, rather crude set-up: Sunlight providing the light source, a telescope collecting the scatter from the sample, and Raman’s eyes serving as the detector. Complementary coloured glass filters were used to select the colour of light striking the sample and to block out the light that was elastically scattered (of same wavelength as the incident light) from the sample.

When light is transmitted through matter, it is scattered in random directions. Most of the scattered light is of the same wavelength (or frequency) as the incident light. This is known as "Rayleigh scattering" and is considered elastic as the scattered photon has the same energy as the incident photon. Some light is inelastically scattered—scattered photon having higher or lower energy than incident one—at a different wavelength (or frequency). This is called "Raman Scattering". This shifting of frequency is also called "Raman Effect", which arises when the incident light excites molecules in the sample, which subsequently scatters the light.

Typically, only one part in a thousand of the total intensity of incident light is Rayleigh scattered, while for Raman scattering this value drops to one part in a million (0.0001 percent). Thus, Raman effect being quite weak, it is a major challenge in Raman spectroscopy to attenuate the light thatis elastically scattered in order to detect the inelastically scattered Raman light.

"Raman Scattering" results from the molecule changing its molecular motions. The energy difference between the incident light and the Raman scattered light is equal to the energy involved in changing the molecule’s vibrational state, and is called the "Raman shift". If the initial beam is sufficiently intense and monochromatic, a threshold can be reached beyond which light at the Raman frequencies is amplified, builds up strongly, and generally exhibits the characteristics of stimulated emission. This is called the stimulated or coherent Raman Effect. A device illustrating the stimulated Raman Effect is sometimes called a Raman laser.

--------------------------------------------------------------------------------------------------------------------------------------------

Niraj K Gupta, from my cell, Voice and Data, September 2000.