Efficient dispersion modeling in optical multimode fiber

Dispersion remains an enduring challenge for the characterization of wavelength-dependent transmission through optical multimode fiber (MMF). Beyond a small spectral correlation width, a change in wavelength elicits a seemingly independent distribution of the transmitted field. Here we report on a parametric dispersion model that describes mode mixing in MMF as an exponential map and extends the concept of principal modes to describe the fiber’s spectrally resolved transmission matrix (TM). We present computational methods to fit the model to measurements at only a few, judiciously selected, discrete wavelengths. We validate the model in various MMF and demonstrate an accurate estimation of the full TM across a broad spectral bandwidth, approaching the bandwidth of the best-performing principal modes, and exceeding the original spectral correlation width by more than two orders of magnitude. The model allows us to conveniently study the spectral behavior of principal modes, and obviates the need for dense spectral measurements, enabling highly efficient reconstruction of the multispectral TM of MMF. A parametric dispersion model that describes mode mixing in multimode fiber enables calibration of the fiber’s multispectral transmission matrix with significantly fewer measurements than existing approaches.