When mode_coupling is enabled ("on"), the model uses mode groups, with total coupling assumed between modes within a given group. In other words, all modes within a group carry equal amounts of power. This allows dealing with the average power signal of a given mode group, rather than each mode individually. Another assumption is that each mode within a group experiences the same longitudinal phase change and has the same average attenuation coefficient, propagation constant, propagation delay, and mode-coupling coefficient. Coupling between mode groups is calculated via a set of coupled power-propagation equations, in which microbends are assumed to dominate the coupling process, leading to nearest-neighbor coupling1.
The model allows four different versions of the mode coupling coefficient, denoted by types A to D in the parameter window of the MMF model.
In large core fibers, the mode count can be very large. In this case, to save computational resources and simulation time, you can enable super-modes, where instead of retaining spatial information for each individual mode, the model calculates a super-mode for each mode group. For more information, please refer to the detailed model description in the ModeSYS manual.
To analyze the impact of mode coupling, consider a simple setup where a 1-mW CW signal with Gaussian mode profile is launched in to two separate MMFs, each 8-km long, 50-µm core diameter with a parabolic refractive index profile. In one of these two fibers, we ignore mode coupling; in the other, we model the mode coupling effect.