Abstract

Unrepeatered wavelength-division-multiplexed (WDM) transmission links, operating without active in-line elements between transmitter and receiver terminals, present a number of advantages in terms of reduced maintenance costs and increased reliability. These kinds of systems have proven to be ideal for communication with or between islands, as well as with offshore infrastructures such as oil platforms or wind farms. As such, they have garnered much interest in recent years, as continuous technological improvements have allowed for longer spans (up to 90 dB loss) and higher capacities (up to the 10s of Tb/s) [1]. Among other solutions, distributed amplification [2] has been a key technology in enabling such achievements, whether in systems relying on installed legacy fiber or in new installations with novel reduced nonlinearity or low-loss optical fibers. Here, we present a new general, simplified, approach to the optimization of unrepeatered systems using distributed amplification. Our approach is based on the procedure first presented in [3,4] for long-distance transmission systems, adapted to the particular constraints of modern unrepeatered transmission, in which full recovery of signal power at the end of the span is not a requirement, allowing for an extra degree of freedom. The method presented is equally applicable to ASK and PSK formats, as well as to any form of WDM configuration, and employs a multi-level approach to greatly reduce the time devoted to full NLSE simulations, by searching for sets of parameters satisfying optimal trajectories in the system configuration space that satisfy a conditional minimization of nonlinear impairments for given values of the optical signal-to-noise ratio (OSNR). Furthermore, following [4], this approach allows for a fast and accurate estimation of the optimal dispersion pre-compensation for each channel based on the estimation of the nonlinear phase shift (NPS) barycentric, accounting for its dependence on factors such as number of channels, input signal power or pump power distribution. In addition, we will compare the design and performance differences between systems relying on legacy fiber and those using a more modern fiber base.

© 2015 IEEE