Research Of Adaptive Optics For Wave Front Distortion Correction In Free Space Optics Communication

Free space optics communication(FSOC)transmits information using laser in atmosphere and deep space directly.The technique has become a fascinating research topic and has been widely used in the applications of satellite-to-ground laser communication,inter-satellite communication,frontier and coast defense and civil communication service because of the advantages of large bandwidth potentiality,unregulated spectrum,relative low power requirement,ease of redeployment and low cost.FSOC gradually gets attention with the promotion of requirements in the subject of the ‘last mile' communication access.However,since atmospheric turbulence brings phase disturbances along propagation paths that are manifested as decrease of the quality of the beam,intensity fluctuation(scintillation),beam wandering and beam broadening at laser receiver.It leads to the significant decrease of coupling efficiency and energy loss at the receiving terminal,which makes highly excessive bit error rate(BER)and results in bad communication performance.Adaptive optics(AO)is a novel technique to detect and correct the wave front distortion at real time.As an integrated discipline which combines science and engineering,AO system has been taken into use in several areas such as human eye vision improvement,high speed laser transmission and astro observation.At present,AO technique plays more and more important role in the research of wave front distortion in FSOC system.This paper first analyzes the effect of atmospheric turbulence on the laser transmitting and FSOC system,and illustrates the AO system on the aspects of wave front sensing,wave front reconstruction and wave front control.We also setup numeral models of AO system to study the wave front distortion correction.Furthermore,we also construct a simple experimental AO system platform to verify part of our proposed methods.The main work in this paper concentrates on the following aspects:1.We present the significance and background of this research at domestic and foreign institutions on FSOC and AO system.And we analyze the technique questions of the traditional AO and sensor-less AO system by combining the development situation.2.We briefly analyze the effect on FSOC link of atmosphere turbulence on the aspects of BER,coupling efficiency and Strehl ratio.We also introduce the fundamental theories of traditional AO and sensor-less AO system.3.To conduct wave front distortion correction based on the traditional AO system,we take time delays in consideration and introduce Kalman filter into AO system as controller for the corrections.Then we present some transformations and optimizations to the Kalman filter based controller to design the fading-memory Kalman filter which makes the system recover quickly to the right status and work normally under the conditions of unexpected interferences that may result in an incorrect estimation.The modified controller can significantly avoid the information loss in the process of ‘self-healing'.At the meantime,we also propose twoequivalent transformations to obtain our sequence Kalman filter and U-D Kalman filter for easy implementation in embedded systems.Furthermore,we develop several swarm intellectual based PI tunning algorithms for the most widely used PI control.4.Since too many sub-apertures in wave front sensor may lead to energy dispersion,the energy would be too low in each sub-aperture for detection,which will lead to abnormal work of AO system,or even FSOC link will not work.Therefore,the research of how to solve this problem in sensor-less AO system is necessary,which is including the model-free methods and the model-based methods.Precisely,we study the control of model-based methods in this paper.We first introduce some previous methods,such as the method of Martin J.Booth,the method of Huang Linhai and the method of Jacopo Antonello.Then we propose a novel BP artificial neural network based method to detect and correct the wave front distortion.This method has the advantages of fewer measurement requirement and better system real-time performance compared to the previous methods.However,its drawback is that some off-line operations are needed.We believe that the well-chosen data structure and highly developed hardware in the near future will solve this problem,and the off-line operations do not make impact on the real-time on-line wave front distortion correction.Additionally,we also study the some model-free random search method,such as the most known stochastic parallel gradient descent(SPGD)algorithm for the problem optimization.We propose our random search optimization algorithms to reduce numbers of iteration,avoid local optimization and failure to meet the real-time requirement.These algorithms are the simulated annealing(SA)algorithm,global optimized algorithm to find global optimized solution with short searching time and modified shuffled frog leaping(MSFL)algorithm which belongs to the swarm intelligence optimized algorithms.They have the advantages of robust searching and global optimization.However,multiple AO systems parallel operation is needed.If one wants to make the most use of the algorithms,the cost of hardware device will increase.However we use serial iterative search to replace parallel searching for the correction of ‘one shoot' static wave front distortion.The Tabu search(TS)algorithm is adopted to map voltage vector space to Zernike coefficient vector space,in order to reduce the dimension of searching space and accelerate searching velocity.The result of the correction by this method shows lower device cost than that of MSFL methods.5.We present our future work on the improvement of FSOC link experimental platform for deep study of FSOC link communication performance.This paper gives detail theoretical research and related numeral simulation and experiment results.The findings indicate that the terminal coupling efficiency increases depends on the performance of AO system,which will improve communication performance of FSOC systems.