Abstract

Ultrashort pulse propagation in optical fiber leads to a wide range of complex dynamics associated with dispersive and nonlinear interactions [1]. A comparably wide range of dynamics can be seen in modelocked fiber lasers, and the concept of the dissipative soliton has provided a powerful framework within which to interpret the various class of behavior that have been observed. Most studies of fiber lasers have focused on cavity designs that avoid instability and produce highly regular pulse trains, but there has also been extensive interest in the highly unstable “noise-like pulse” (NLP) regime, where a large number of ultrafast pulses evolve randomly [2]. With highly nonlinear fiber in the cavity, NLP lasers can generate output spectra spanning 100’s of nm, but achieving these broadest bandwidths requires time-consuming optimization of the high dimensional parameter space associated with the nonlinear polarization rotation based saturable absorber (SA). In this paper, we replace this trial and error approach with a genetic algorithm (GA) that auto-aligns the laser into the broadband NLP regime.

© 2023 IEEE