Algorithms that correct for volume atmospheric turbulence in coherent imagery are computationally intensive, typically requiring several iterations to converge to a solution with a split-step model, where each iteration involves multiple optical propagation computations. We examine the sampling requirements for split-step modeling using phase-space optics and show that we can propagate fields accurately, using array sizes that are 2-4× smaller than the array sizes used in a typical split-step model. These smaller array sizes can be used when the aperture and field stops for the imaging system are used as intermediary planes for individual propagation steps. We evaluate the fidelity of vacuum split-step propagation results, describe split-step model adjustments needed to accommodate diffraction and turbulence effects, and illustrate how we use split-step models for analyzing the expected performance of turbulence-compensation algorithms.
|