Advanced S$^2$ Imaging: Application of Multivariate Statistical Analysis to Spatially and Spectrally Resolved Datasets

Spatially and spectrally resolved imaging has become an important tool for multimode fiber characterization, both in the realms of group index dispersion measurement and modal geometry analysis. However, limited resolution in group delay space, distributed scattering and noise sometimes make the task of identifying meaningful modal interference figures difficult. In this paper, an application of multivariate statistical analysis to S
²
datasets that effectively isolates mode pair interferences, both spectrally and spatially, is presented. A method to use those components to retrieve the power in each mode is also shown.