Fiber Nonlinearity Compensation: Commercial Applications and Complexity Analysis

Fiber nonlinearities define the ultimate performance bound for optical communication systems. Todays 100 Gb/s commercial products employ advanced digital signal processing (DSP) algorithms, capable to adequately address linear channel impairments, leaving fiber nonlinearity compensation (NLC) as the next logical step to improve transmission performance, and consequently diminish the need for signal regeneration. Over the last few decades, several techniques have been presented to minimize or mitigate channel nonlinearities, ranging from specialized link designs in direct-detect legacy networks to advanced DSP algorithms in coherent systems. However, specifically in the coherent age, NLC has always been perceived as an extremely complex approach, allowing insignificant gains when practical implementations are considered. In this paper, we focus on the real-world commercial use cases and complexity tradeoffs for NLC, and review several application scenarios, including homogeneous and heterogeneous networks, dispersion unmanaged and dispersion managed link infrastructures, flex-grid networks, and short reach to ultra long-haul transmission applications. We establish that in various practical use cases, NLC may enable substantial performance gains, well beyond conventionally acknowledged bounds.