Temporal Energy Analysis of Symbol Sequences for Fiber Nonlinear Interference Modelling via Energy Dispersion Index

Exponentially-Weighted Energy Dispersion Index for the Nonlinear Interference Analysis of Finite-Blocklength Shaping

“…Similar to our previous work [14], we study how well EEDI can predict the blocklength-dependent effective SNR. A single-polarization multi-span wavelength division multiplexing (WDM) system is simulated by using the split-step Fourier method.…”

“…To the left of the vertical dotted line, the effective SNR decreases significantly as blocklength n increases. We call this the blocklength-dependent region [14]. Then, the effective SNR decreases slowly until it reaches a floor.…”

“…In our previous study, we assumed that the NLI generation is dominated by symbol energies within a finite time window [14]. In this paper, we extend this time window to infinity and we introduce a decay factor to weigh the effect of temporal separation between symbols on the NLI generation.…”

“…Recently, in [14] we analyzed the statistical properties of symbols generated by constant composition distribution matching (CCDM) [15] with finite blocklengths. Inspired by the behavior of time-domain first-order perturbation NLI models [16] and the finite memory Gaussian noise (GN) model [17], we found that the variations of the windowed symbol energy are crucial for the NLI generation.…”

“…Inspired by the behavior of time-domain first-order perturbation NLI models [16] and the finite memory Gaussian noise (GN) model [17], we found that the variations of the windowed symbol energy are crucial for the NLI generation. We also proposed in [14] a precise metric to quantify the effect of energy variations Corresponding author: K. Wu (email: k.wu@tue.nl).…”