Efficient distortion metrics calculation for hyperspectral waveforms in broadband nonlinear amplifiers with memory

We present a method for efficient, physics-based calculation of distortion metrics for hyperspectral waveforms (i.e. wide-band, with large number of frequency components) in broadband nonlinear amplifiers with memory. It leverages the acceleration potential of our recently developed framework for massively parallel frequency-domain simulation of hyperspectral waveforms [1]. We introduce a generalized distortion metric that accounts for memory effects, such as dispersion and wave reflections, and we study the effect of varying the parallelization parameters on the accuracy of the distortion calculation.