ITU group drafts specs for WDM and optical amplifiers
LOUISE ALTERMAN
Experts on Synchronous Digital Hierarchy optical line systems met this past February in Richardson, TX, to continue work on emerging Recommendations for wavelength-division multiplexing (WDM) and optical amplifiers. The group, officially known as the International Telecommunication Union`s (ITU) Study Group 15 Question 25 experts, is led by Rapporteur John Eaves of Bellcore.
One draft document under progress is Recommendation G.mcs, "Optical Interfaces for Multichannel Systems with Optical Amplifiers," which pertains to WDM systems operating in the 1550-nanometer window. The document covers systems of up to four and eight channels--operating at rates of up to 2.5 gigabits per second per channel--selected from a grid of frequencies referenced to 193.1 terahert¥and spaced every 100 gigahertz. The group realizes that this grid does not guarantee transverse compatibility because more than 40 allowed channels are available, from which a maximum of eight channels must be chosen. However, because no agreement was reached on which channels to choose, the compatibility issue will be covered in the next version of the document.
Power-level issues remain
Some remaining issues in the development of G.mcs include the specifications for power levels (per channel and total), and the use of these power levels to develop optical-link budgets. These specifications have proven more complicated than those for single-channel systems because the total power may not be divided equally among the channels, and the amplified spontaneous emission noise of the optical amplifiers can limit the span lengths.
The group tentatively agreed on a methodology for calculating these power levels. This methodology is slated for use in single-span systems and in systems with multiple spans between inline optical amplifiers. Span lengths are targeted at 80 and 120 kilometers, with maximum attenuation values of 22 and 33 decibels, respectively.
The methodology is based on setting upper and lower limits on the power levels at the output of the optical amplifiers. It calculates the signal-to-noise ratio (SNR) of the channels as a function of the per-channel output power, number of spans (between optical amplifiers) and the optical loss per span. It is then possible to determine the minimum per-channel power required to achieve an adequate SNR. The SNR is correlated with the bit-error rate, but the exact SNR required to achieve a 10-12 bit-error rate is still under discussion. The maximum per-channel power level will be determined by nonlinearities and the maximum total output power, which is set at +17 dBm because of safety considerations. After these power limits are determined, margins will be added to account for temperature, aging and gain variations among different channels.
After all these methods are agreed upon, and power levels--including margins--are determined, the experts will calculate how many spans of 80 and 120 km are practical. At least one delegate has indicated a maximum of three spans of 120 km. q
Louise Alterman is a member of technical staff, lightwave systems engineering group, at Lucent Technologies, Bell Laboratories, in North Andover, MA, and is editor of Recommendation G.mcs. She holds a B.S. from Cornell University and an M.S. from the Massachusetts Institute of Technology, both in electrical engineering.