Structured light entities, chaos and nonlocal maps
A. Yu. Okulov
TL;DR
The paper develops nonlocal iterative maps with Green-function kernels $K(\vec{r}-\vec{r}')$ as an efficient numerical framework that is formally equivalent to nonlinear parabolic PDEs of the Ginzburg-Landau type, enabling accurate modeling of spatial chaos, solitons, vortices, and vortex lattices in optical resonators under multimode noise. By expressing propagation as convolution maps and relating them to slice-by-slice evolutions, it links discrete map dynamics to evolution equations such as diffusion, diffraction, and NLS-GLE, with explicit connections to the KPP equation and the Gross-Pitaevskii framework. The work demonstrates the emergence of localized wavetrains and solitons as fixed points of nonlocal maps, and shows how periodic transverse structures and vortex lattices arise in cavities with varying Fresnel numbers, including Talbot self-imaging phenomena and phase-locked lattices. Incorporating multimode noise yields realistic relaxation oscillations and spectra, illustrating the method’s ability to reproduce experimental patterns while maintaining rapid convergence via FFT-based computation. Overall, the nonlocal-map approach provides a flexible, boundary-aware, computationally efficient bridge between discrete dynamical systems and continuous PDE descriptions of pattern formation in nonlinear optical media.
Abstract
Spatial chaos as a phenomenon of ultimate complexity requires the efficient numerical algorithms. For this purpose iterative low-dimensional maps have demonstrated high efficiency. Natural generalization of Feigenbaum and Ikeda maps may include convolution integrals with kernel in a form of Green function of a relevant linear physical system. It is shown that such iterative $nonlocal$ $nonlinear$ $maps$ are equivalent to ubiquitous class of nonlinear partial differential equations of Ginzburg-Landau type. With a Green functions relevant to generic optical resonators these $nonlocal$ $maps$ emulate the basic spatiotemporal phenomena as spatial solitons, vortex eigenmodes breathing via relaxation oscillations mediated by noise, vortex-vortex and vortex-antivortex lattices with periodic location of vortex cores. The smooth multimode noise addition facilitates the selection of stable entities and elimination of numerical artifacts.
