

The accuracy of wavefront reconstruction from discrete slope measurements in adaptive optics (AO) systems depends on the sampling geometry, coherent length of the incoming wavefront, wavefront sensor specification and accuracy of the reconstruction algorithm. In this paper, simulation of Gaussian beam propagation in a turbulent atmosphere was carried out and the slope measured by Shack- Hartmann wavefront sensor (SHWS) was simulated. Wavefront reconstruction based on iterative discrete Fourier transforms was simulated, and then reconstructions with Fried and Hudgin inverse filter in spatial frequency domain were compared with Roddier iterative Fourier reconstruction. In these algorithms, no spatial boundary control is required. Our simulations showed that Hudgin algorithm results in a bigger Strehl ratio, so its reconstruction performance is more accurate than Fried and Roddier algorithms. In comparison to vector matrix multiplication reconstruction, this algorithm is faster and more accurate.