On The Methods Of Fiber Nonlinearity Estimation Within Frequency Domain

Ultra-high-speed long-haul optical transmission systems, the commanding heights of the next-generation optical communications technology, are attracting the most interests of researchers all-over the world. To overcome transmission loss and have adequate OSNR (Optical Signal Noise Ratio), ultra-long haul optical transmission requires a high incident power into the fiber, which easily lead to a nonlinear optical effect. So nonlinear noise will be the last barrier to the development of ultra-high-speed long-haul optical transmission systems. Nowadays, people prefer to use Back-propagation (BP) method, which is very time consuming, to compensate the effect of the nonlinear noise. If a closed-form expression of the fiber nonlinear effect is available and accurate enough through which we can figure out the value of the fiber nonlinear noise given some key system parameters that will definitely promote the process of nonlinear noise compensation and the design of the ultra-high speed optical transmission systems. This paper studies the frequency domain estimation methods of optical fiber nonlinear effects, and the research content is divided into two parts:the first part of the study is based on PM (polarization multiplexed) CO-OFDM simulation platform (herein after referred to as CO-OFDM systems), under different optical communications system link parameters and the different design parameters. The second part is based on a PM Nyquist single carrier system (herein after referred to as Nyquist system) at various optical communication system parameters. Through in-depth research we get the following two conclusions: First, frequency domain nonlinear effects estimation method based on CO-OFDM system has some limitations that it does not apply to high-level modulation formats such as16-QAM. The reason is nonlinear noise has been modeled as simple additive noise in this kind of frequency-domain nonlinear effects estimation method, but obviously nonlinear phase noise which should also be included is neglected. We show for the first time through comprehensive simulation under both uncompensated transmission (UT) and dispersion managed transmission (DMT) systems that the statistical distribution of the nonlinear noise within the polarization multiplexed16-state quadrature amplitude modulation (PM-16QAM) Coherent Optical OFDM (CO-OFDM) system deviates from Gaussian distribution in the absence of amplified spontaneous emission (ASE) noise. We also observe that the dependences of the variance of the NLI noise on both the launch power and the transmission distance (logarithm) seems to be in a simple linear way. Second, through the study based on the Nyquist single carrier systems, we find that Brute-Force Integration method is only suitable for signal after single span transmission where the estimation results are accurate enough. But for signal transmitted through multiple spans, Brute-Force Integration method for nonlinear noise estimation is no longer applicable. More appropriate method is to regard the power spectral density of the transmitted signal as constant value, and only estimate the nonlinear noise power spectral density at the center frequency. By applying this method, for all modulation formats and dispersion compensation ratio, we see good agreement between the theoretical estimated and simulation values.In summary, this paper carried out a detailed study of nonlinear effects estimation method within two super-channel technology CO-OFDM and Nyquist, major innovations and achievements are as follows:1. A detailed study of nonlinearity estimation method based on CO-OFDM system simulation platform frequency optical communication systems at different link parameters, different CO-OFDM design parameters demonstrated closed-form expression model of nonlinearity estimation does not apply to16QAM and higher order modulation formats.2. Through deep analysis of the nonlinear noise distribution model within high-speed16QAM modulated transmission systems, we find its distribution deviates from the Gaussian distribution. This conclusion not only can help to explain the inaccuracies of the existing nonlinear models, but also contributes to establish a more comprehensive nonlinearity estimation model.3. A thorough study of the nonlinearity estimation method based on Nyquist single carrier systems under different link parameters, different design parameters shows that the Brute Force Integration method does not apply to case of multi-span transmission.