Translator Disclaimer
19 February 2009 Improved neural network algorithm: application in the compensation of wavefront distortion
Author Affiliations +
A Free Space Optical Communication (FSO) system transmits modulated light through atmospheric media. Because of the uneven distribution of refractive index result from atmospheric turbulence, the phase distribution of light is changed leading to distortion of wavefront and requiring reconstruction at the receiver. However, current wavefront compensation relies on channel modeling which has difficulties in extracting channel information from highly random turbulent atmosphere. In this paper, a wavefront reconstruction system based on neural network algorithm is constructed. The neural network requires little channel information but predicts distortion by past experience. Then, distorted phase distribution is adaptively revised when light passes through a piezoelectric ceramic deformable mirror controlled by neural network. Dynamic study factors are added to neural network algorithm as improvement which adjusts the study speed of the system according to turbulence intensity providing best result between respond time and reconstruction accuracy. In addition, light transmitted in atmospheric channel is studied.
© (2009) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Zhou Zhou, Xiuhua Yuan, and Jin Wang "Improved neural network algorithm: application in the compensation of wavefront distortion", Proc. SPIE 7279, Photonics and Optoelectronics Meetings (POEM) 2008: Optoelectronic Devices and Integration, 72791F (19 February 2009);

Back to Top