Estimating the atmospheric correlation length with stochastic parallel gradient descent algorithm

Abstract

The atmospheric turbulence measurement has received much attention in various fields due to its effects on wave propagation. One of the interesting parameters for characterization of the atmospheric turbulence is the Fried parameter or the atmospheric correlation length. We numerically investigate the feasibility of estimating the Fried parameter using a simple and low-cost system based on the stochastic parallel gradient descent (SPGD) algorithm without the need for wavefront sensing. We simulate the atmospheric turbulence using Zernike polynomials and employ a wavefront sensor-less adaptive optics system based on the SPGD algorithm and report the estimated Fried parameter after compensating for atmospheric-turbulence-induced phase distortions. Several simulations for different atmospheric turbulence strengths are presented to validate the proposed method.

Phase reconstruction of two interfering fields using a lenslet array and an optimization method

“…The method of minimizing the cost function is the SPGD algorithm which has been successfully used in phase reconstruction [10,21] and phase distortion correction [14,15,26,27] for its high convergence speed. The phase difference distribution Δfat the reconstruction plane is derived using the phase shifting technique.…”