Analysis Of Four-wave-mixing Noise In Dispersion Managed DWDM System

With the development of Dense Wavelength Division Multiplexing (DWDM) technology,increasing of optical fiber transmission channel number, reducing of channel spacing and increasingof launch power, nonlinearities in optical fiber, including four-wave mixing (FWM), have moreeffect on transmission performance. Suppression methods of FWM effect are mainlydispersion-managed technology and adopting new modulation format.This paper focuses on studying FWM in dispersion management DWDM systems with twokinds of optical modulation formats. The main innovations are given as follows:1. Based on finite-bandwidth stationary random noise spectral analysis theory, the calculationmodel of standard deviations of degenerate and non-degenerate four-wave mixing noise indispersion-managed DWDM systems with frequency shift keying (FSK) format is derived. Thecalculation model includes random bit sequences, inter-channel walk-off effect, oscillation of pulsewidth with transmission distance, and stochastic variation of birefringence strength and orientationalong fiber. In the paper, a concept of FQP is defined. The results of example calculation show that:when local dispersion is small and transmission distance is long, values of FQP are larger, asindicate the impact of FWM noise on performance of transmission system is larger. In the case ofallowed maximum input powers for the system, variances of overall FWM noise are observed to beabout half of those of signal-ASE beat noise. When the local dispersion value and the transmissiondistance are large enough, compared with the degenerate FWM noise, non-degenerate FWM noisecan be negligible.2. This paper derives a calculation model of standard deviations of degenerate four-wave mixingnoises in dispersion managed DWDM system with differential phase shift keying (DPSK) format.Compared with on-off keying (OOK) format, the standard deviations of FWM noise in case ofDPSK format are smaller for same average launched power. Bit rate per channel has little effect onthe standard deviation of FWM noise. The average power has great influence on the standarddeviation of FWM noise.3. This paper derives a calculation model of non-degenerate four-wave mixing noise indispersion managed DWDM system with DPSK format. The model also includes some randomfactors, but ignores oscillation of pulse width with transmission distance. Taking into account thenon-degenerate and degenerate four-wave mixing noise and other noise sources simultaneously, transmission performance of such system is analyzed numerically. The results indicate that: theimpacts of local dispersion and launch power on Q-factor and FQP are same as the case of FSKformat. When local dispersion increases, decreasing rate of FQP and increasing rate of Q-factor isbecome small.