The Application And Modification Of Split-Step Fourier Method In The Optical Fiber Communication Systems Simulation

The nonlinear Schrodinger equation (NLSE) is quite useful in the optical communication field, and has been applied widely to the optical communication systems simulation. Numerical calculation methods are usually more widely applied in NLSE because of its complexity. The most commonly used algorithm is the split-step Fourier method (SSFM).In simulations of optical communication systems, the time-window and step-size should be selected correctly. The small initial time-window will result in the distortion of the numerical calculation in the situation of the pulse energy diffusing fast, because it makes the pulse energy quickly diffuse to the edge of time-window in a short time. On the other hand, for the sake of obtaining the exact numerical result, the step-size tends to zero in the constant step-size method. But the smaller is the step-size, the higher is the requested computational time. Besides, the improper step-size selection can lead to spurious FWM tones.In this paper, by monitoring tne pulse energy, we modify the SSFM to achieve the adaptable control of the time-window. When the pulse energy is going to diffuse to the edge of time-window, the modified method can automatically extend the time-window to avoid the distortion. As to the step-size selection, we implement the local-error method to improve the efficiency and accuracy of the numerical calculation.We apply this modified method to some single-channel systems simulation. In the cases of dispersion broadening, (?)hird-order dispersion and couple of solitons, we can obtain the exact result of numerical calculation and improve the efficiency of simulation. In multi-channel simulations, we use the adaptable method of the time-window to analyze the XPM exactly. On the other hand, in order to resolve the spurious FWM tones which result from the inaccurate step-size distribution, we analyze and compare the reliability and efficiency of constant step-size method, logarithmic step-size met(?)od and modified constant step-size method.