Overcoming degradation in spatial multiplexing systems with stochastic nonlinear impairments

Abstract

Single-mode optical fibres now underpin telecommunication systems and have allowed continuous increases in traffic volume and bandwidth demand whilst simultaneously reducing cost- and energy-per-bit over the last 40 years. However, it is now recognised that such systems are rapidly approaching the limits imposed by the nonlinear Kerr effect. To address this, recent research has been carried out into mitigating Kerr nonlinearities to increase the nonlinear threshold and into spatial multiplexing to offer additional spatial pathways. However, given the complexity associated with nonlinear transmission in spatial multiplexed systems subject to random inter-spatial-path nonlinearities it is widely believed that these technologies are mutually exclusive. By investigating the linear and nonlinear crosstalk in few-mode fibres based optical communications, we numerically demonstrate, for the first time, that even in the presence of significant random mixing of signals, substantial performance benefits are possible. To achieve this, the impact of linear mixing on the Kerr nonlinearities should be taken into account using different compensation strategies for different linear mixing regimes. For the optical communication systems studied, we demonstrate that the performance may be more than doubled with the appropriate selection of compensation method for fibre characteristics which match those presented in the literature.

On the Performance of Digital Back Propagation in Spatial Multiplexing Systems

“…Section IV presents and analyzes simulation results for the different transmission models in section II over a wide range of DMD and XT values; extending our previous results by considering a larger number of WDM channels over span lengths optimized to minimize power consumption [39,40]. Section V extends our recent results on the application of DBP to WDM-SDM systems [41][42][43] by: (i) analyzing the dependency on the number of back-propagated channels; (ii) considering a wider range of LMC and DMD scenarios; (iii) extending the characterization of the GD spreading in the strong LMC regime. Section VI draws the final conclusions.…”

Digital back propagation performance in spatial multiplexing systems

“…Please see [33] for uncoupled NL coefficients. When considering different DMD values, we simply scale the GD vector in Table 1 instead of re-optimising the profile to avoid changing other fibre characteristics (as in [18]). For full details on the simulation setup, channel estimation and equalization please see [18].…”

“…When considering different DMD values, we simply scale the GD vector in Table 1 instead of re-optimising the profile to avoid changing other fibre characteristics (as in [18]). For full details on the simulation setup, channel estimation and equalization please see [18]. Finally, after transmission, the Q-factor of the centre channel is estimated using the mean and standard deviation of the received symbols [34].…”

“…Here we demonstrate the applicability of digital back propagation (DBP) to address the IM-NL penalties in SDM systems. After reviewing the models proposed to transmission in SDM fibres, we review our recent work [18] on simplified DBP methods for the different operational regimes determined by DMD and LMC.…”

Digital Back Propagation via Sub-Band Processing in Spatial Multiplexing Systems

“…systems has been under study [36,37] having been shown that their impact is non-negligible. As a consequence, the investigation of applying digital back propagation (DBP) to address the IM-NL penalties in SDM systems has attracted interest [38][39][40][41] showing that considerable improvements can be achieved for weak to intermediate linear coupling regime [39] and identifying cases where increasing DBP bandwidth towards full field compensation degrades the performance [41]. In this paper, we explore alternative ways of implementing the nonlinear compensation step in the split-step Fourier method (SSFM) for SDM systems which are characterized by significant decorrelation (as induced by DMD and LMC) between the spatial-and wavelength-division-multiplexed (WDM) channels.…”