Evaluation of Nonlinear Interference Effects in a Dispersion Managed(DM) Optical Fiber: Performance and Transmission Analysis of 16QAM Modulation using Split Step Fourier Method Reinhardt Rading Department of Engineering and Architecture University of Parma Parma, Italy Abstract—This paper investigates the impact on the optical signal-to-noise ratio (OSNR) of the residual per span (RDPS) in a N × 100km dispersion managed system with zero total accumulated dispersion from input to output using split step Fourier method (SSFM) -Monte Carlo simulation. This paper will show that the nonlinear interference NLI does in-fact impact the performance yielding different best working power depending on the value of Nx100 km span and the type of dispersion managed link. The paper will show that dispersion uncompensated optical links are preferable to dispersion managed fibers in equalizing NLI effects in long haul optical links. Index Terms—Optical Signal-to-Noise Ratio(OSNR), Disper- sion Managed(DM), Residual per Span(RDPS), Nonlinear Inter- ference (NLI) I. I NTRODUCTION Due to huge data demand and the need for efficient spectrum usage, coherent systems are becoming the norm for long haul optical transmission systems and network. After several of kilometers of transmission, the optical system experiences signal degradation resulting from the noise produced in the system. One metric used to characterize degradation and measure the quality of transmission (QoT) is the use of optical signal-to-noise Ratio (OSNR)- defined in optical domain as the ratio of optical signal power to noise power. Different models [1] have been suggested including for example,the Gaussian Noise (GN) model [2], used to predict performance of an optical link in the nonlinear regime. Apart from the amplified spontaneous (ASE) noise, nonlinear signal interference (NLI)-a noise like power induced by nonlinear effects -is factor that contributes to signal degradation in long haul optical system. Many models assumes that NLI is a small perturbation of nonlinear Schrodinger equation [3] but it has an immense effect on nonlinear propagation and thus should be included in QoT prediction. The aim of this paper is to show that NLI does in-fact impact the performance,yielding different working power depending on distance and the dispersion on the link. The paper is divided into five sections: section II discusses the theoretical part, section III explains the numerical analysis, section IV states the simulation results, and section V discusses the simulation results. The paper finally draws conclusion in section VI. II. THEORY An optical transmission system traversing several of kilo- meters will experience signal impairments resulting from ASE noise (results from signal amplification using Erbium doped fiber amplifiers) and nonlinear noise (results from nonlinear fiber response to the input power). Signal distortion due to nonlinear fiber response increases with an increase in input power, increase in transmission length, and transmission rate. High power in optical systems results to a change in the material refractive index with respect to the electric field intensity leading to a phenomenon known as Kerr Effect [4]. The Kerr Effects leads to signal distortion and limits the distance [5] the signal can travel without experiencing distortion. For an N span system, the total NLI is the sum of N noise random variables. Gaussian model assumes that the noise contributions are uncorrelated and thus ease of use and to determine their influence. The limitation of Gaussian model is that the above assumption only holds over in a dispersion uncompensated optical links. This paper focuses on dispersion managed links with disper- sion compensating fibers (DCF) being used to compensate for dispersion in the link and are applied pre-transmission, inline or post transmission. Different dispersion managed (DM) links exists with a 0 residual dispersion per span (0 RDPS) or a 30 residual dispersion per span (30 RDPS). Non-linear effects in a fiber can be determined by solving the Non-Linear Schrondiger Equation given by equation 1 using split step Fourier method (SSFM). Compared to other NLSE models, SSFM is two magnitude orders faster [6]. dA dZ = − α 2 A − B 1 dA dt + j β 2 2 d 2 A dt 2 + β 3 6 d 3 A dt 3 − jγ |A| 2 A (1)