Bit Pattern Length Dependence of Intrachannel Nonlinearities in Pseudolinear Transmission

Wickham, Lisa K.; Essiambre, René-Jean; Gnauck, A.H.; Winzer, Peter J.; Chraplyvy, A. R.

doi:10.1109/lpt.2004.826782

Cited by 63 publications

(35 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…From (17), we get a real kernel in the frequency domain by imposing the condition (18) Condition (18) cannot be met by all pairs . The best that can be done is equalizing the first-order Taylor expansion of the arctan term, i.e., , obtaining an optimal precompensation equal to (19) Equation (19) is the straight-line rule (SLR) mentioned in the introduction for a terrestrial DM system and supports the empirical rule found in [15] and explained in [16] with a simple phenomenological model. It also has the same functional behavior as the rule proposed in [13] based on minimization of intrachannel cross-phase modulation, except for a factor multiplying in the SLR in [13], which is not needed to best match the optimal precompensation for NRZ-OOK found by both simulations and experiments.…”

Section: Large-signal Perturbative Analysissupporting

confidence: 56%

“…Knowledge of such a memory is mandatory to run accurate numerical simulations and laboratory measurements with a sufficiently long pseudorandom bit sequence to avoid patterning effects. We discuss the dependence of memory on DM parameters and show that, in DM-optimized systems, the memory scales linearly with bit-rate instead of the quadratic growth found in GVD-dominated systems [19]. We finally discuss the impact of worst case bit-patterns on system performance and clarify that the well-known resonance of eye closure penalty versus map strength found through simulation [9]- [11] is indeed an artifact due to insufficient length of the employed PRBS.…”

Section: Introductionmentioning

confidence: 85%

“…Moreover, we will prove the following novel result. v) The minimum pseudorandom binary sequence (PRBS) length needed to run simulations that correctly account for intersymbol interference scales linearly with bit-rate in optimized periodic DM systems [18], unlike the quadratic growth found in the pseudolinear case dominated by GVD [19]. In this paper, it is shown that the DM-NLSE kernel contains most of the needed information for system optimization.…”

Section: Introductionmentioning

confidence: 93%

“…It is simple to show that our SLR corresponds to choosing so as to set the average value of such a pdf to zero, while Killey's formula (for a single DM span) corresponds to setting the median of to zero. We will numerically verify the accuracy of the SLR (19) in Section VI-A. Such optimal precompensation was derived in the assumption of zero residual GVD and amounts to creating an optical link as close as possible to one with only self-phase modulation (SPM), for which it is exactly .…”

Section: Large-signal Perturbative Analysismentioning

confidence: 99%

See 3 more Smart Citations

A Unified Design Framework for Single-Channel Dispersion-Managed Terrestrial Systems

Bononi

Serena

Orlandini

2008

J. Lightwave Technol.

View full text Add to dashboard Cite

Abstract-This paper provides a unified framework to the design, performance optimization, and accurate numerical simulation of periodic, dispersion-managed (DM) single-channel long-haul optical transmission systems for nonsoliton on-off keying (OOK) modulation. The focus is on DM terrestrial systems, with identical spans composed of a long transmission fiber compensated at the span end by a linear dispersion compensating module, with pre-and postcompensation fibers at the beginning and end of the link. The framework is based on the dispersion-managed nonlinear Schrödinger equation (DM-NLSE). First, expressions of the DM-NLSE kernel are provided both in the frequency and the time domain, and a novel map strength parameter, appropriate for terrestrial systems, is introduced. It is then shown that the DM-NLSE contains all the basic information needed for system design, as summarized by three parameters: i) nonlinear phase, ii) in-line dispersion, and iii) map strength. Through a large-signal perturbative analysis of the DM-NLSE, the well-known linear relationship between the in-line dispersion and the optimal precompensation is derived, along with the large-signal step response of the DM link, from which the ghost pulses energy growth and a first estimation of the link memory are derived. The DM-NLSE is then linearized around the average signal field to get the amplitude/phase small-signal system matrix of the overall DM link, including pre-and postcompensation. By a singular-value decomposition of the small-signal DM link matrix, a novel expression of the memory of the optimized DM link is finally provided. Knowledge of such a memory is mandatory to run accurate numerical simulations and laboratory measurements with a sufficiently long pseudorandom bit sequence to avoid patterning effects.Index Terms-Dispersion-managed nonlinear Schrödinger equation, optical fiber memory, perturbation methods, pseudorandom binary sequence (PRBS) length, small-signal analysis.

show abstract

Section: Large-signal Perturbative Analysissupporting

confidence: 56%

Section: Introductionmentioning

confidence: 85%

Section: Introductionmentioning

confidence: 93%

Section: Large-signal Perturbative Analysismentioning

confidence: 99%

See 2 more Smart Citations

A Unified Design Framework for Single-Channel Dispersion-Managed Terrestrial Systems

Bononi

Serena

Orlandini

2008

J. Lightwave Technol.

View full text Add to dashboard Cite

show abstract

“…For -R s 2 β 2 /α > 0.2 dispersion leads to overlapping of an increasing number of pulses and intrachannel nonlinear effects become the dominating nonlinear impairments. A major obstacle for obtaining accurate results in the presence of intrachannel nonlinear effects (especially for -R s 2 β 2 /α > 1) is the large system memory and therewith needed long bit sequences [8]. In our simulations 1024 symbols are considered.…”

Section: System Setupmentioning

confidence: 99%