NTT extends praseodymium fiber amplifier into two stages
PAUL MORTENSEN
Nippon Telegraph and Telephone Corp. in Tokyo has developed an optical-fiber amplifier that offers high output at low noise in the 1300-nanometer band for use in short-haul lightwave communications networks. The amplifier development is part of an ongoing research effort for the proposed nationwide fiber-to-the-home project scheduled for completion by the year 2010.
For fiber-optic research and other telecommunications applications, NTT is currently spending 290 billion yen (approximately $3.4 billion) annually, which is the highest budget in Japan`s telecommunications industry.
Previously, an optical-fiber amplifier suitable for 1300-nm optical transmission systems for medium-distance and local transmission was not available. The major obstacle was partly the result of the lack of fiber-optic cable material that offered high efficiency and low loss for amplification. Meanwhile, erbium-doped fiber amplifiers developed in 1989 as optical fiber amplifiers for 1500-nm band operation in long-distance networks have already reached the commercial development stage.
Special fiber
The NTT optical amplifier design uses a special fiber made by doping fluoride glass (zirconium fluoride and barium fluoride) with praseodymium to amplify the optical signal without conversion to an electrical signal. The company has confirmed the amplifier`s performance in both 10-gigabit-per-second digital and 40-channel optical analog video transmission experiments.
An NTT spokesperson notes that the optical amplifier is expected to cut costs by decreasing the size and number of parts implemented in conventional large, expensive repeaters, and by a reduction in the number of amplifiers needed in networks. In addition, the amplifier is predicted to provide advanced video services. NTT is running approximately 20 video transmission trials with government and corporate sponsors, notably trading companies.
In 1990, NTT discovered praseodymium could be used for fiber amplification in the 1300-nm band and began the development of praseodymium-doped fiber amplifiers. The company also began to study compositions for fluoride glass in fiber. After modifying the fabrication process, such as the refining method for raw materials and fabrication temperature, the company designed a low-loss fiber for optical amplification.
One of the problems encountered involved noise, which is caused by the reflection of light within the amplifier. To achieve high output at low noise, NTT adopted a two-stage amplifier configuration. The structure consists of a low-noise amplifier unit using a fluoride fiber that minimizes light scattering (first stage) and a high-output amplifier (second stage).
Another source of reflection within the amplifier is the connection between the fluo ride fiber and the silica fiber. By making the endfaces of the fibers slant at an angle of 25 degrees, NTT cut the reflection at the connection point to less than 1/300 of conventional levels. The result is a high-performance praseodymium-doped fiber amplifier operating at 1.3 microns with a signal output power of 20 decibels relative to milliwatts and 6-dB noise.
In 10-Gbit/sec digital transmission tests, the new optical amplifier technology extended the transmission distance from 70 kilometers to 110.8 km. The 40-km extension matches the theoretical calculation for an amplifier with an output of 20 dB/km used in 1.3-micron band fiber with an attenuation of 0.5 dB/km.
For early commercialization, NTT plans to subject the 1.3-micron fiber amplifier design to further reliability tests and to optimize the configuration for use with a semiconductor laser, replacing the currently used solid-state laser.
To check long-haul transmission, NTT transmitted 80-Gbit/ sec soliton signals over 500 km for the first time, surpassing the results of conventional linear transmission schemes. According to NTT, there are two demanding applications for soliton communication. One is long-distance transoceanic communications over 10,000 kilometers in which the transmission speed is limited from 5 to 40 Gbits/sec. The second deals with shorter distance communications over 1000 km at a transmission speed of 100 Gbits/sec.
Many groups have undertaken transoceanic experiments using loop circulation or straight-line transmissions. However, few companies have attempted the NTT experiments. The main reason is that the short pulsewidth of less than 10 picoseconds is not easy to generate with a gain-switched laser diode and spectral filtering. The drawback of such a high-speed soliton system is that the amplifier spacing must be very short because the soliton period is a few tens of kilometers. In addition, fiber dispersion irregularities must be reduced: The standard deviation of the soliton pulses should be much smaller than that for long-haul soliton communications, where long pulses are needed.
In NTT`s experiments, the soliton source used an actively mode-locked erbium fiber-ring laser that emitted 2.7 to 3.0-psec soliton pulses at a repetition rate of 10 gigahertz. The experimenters used dispersion-shifted fibers with an average dispersion of -0.19 psec/km/nm at 1552 nm. Operational success resulted from the regenerative mode-locked optical fiber soliton laser, a stable optical multiplexer utilizing a planar lightwave circuit and an optical demultiplexer containing a nonlinear optical loop mirror. n
Paul Mortensen writes from Tokyo.