| Space laser communication refers to a technology that uses a laser beam as a carrier wave to directly transmit information in space,also known as "free space optical communication." Because of its short wavelength,high transmission rate,good confidentiality and strong anti-interference ability,it has become a research hotspot in recent years.Satellite-to-ground laser communication is another research hotspot after the continuation of inter-satellite laser communication.For satellite-to-ground laser communication links,atmospheric turbulence is the main factor limiting its communication performance.Therefore,the impact of atmospheric turbulence on laser communication needs to be analyzed and suppressed urgently.The turbulence degraded spot image is an important basis for analyzing the turbulence effect,and the satellite-borne laser communication terminal beacon or signal light spot image is the first choice for studying the turbulence of the satellite-to-earth link in this thesis.However,the amount of image data generated by high-resolution beacon photodetectors or high-frame-rate signal photodetectors often reaches the order of Gb/s,which puts great pressure on data buffering and transmission,so it is necessary to develop Data compression work.In the process of data acquisition,due to the accuracy requirements and the Nyquist sampling theorem,it is necessary to acquire images with extremely high resolution.At the same time,in the process of data storage and transmission,in order to reduce the pressure of data transmission and storage,it is necessary to compress high-precision images acquired at a great cost to reduce the amount of data.Compressed sensing technology can directly collect compressed information that meets the requirements of information entropy during the sampling process,thus breaking through the limitations of the Nyquist sampling theorem.Compared with traditional image compression,compressed sensing requires fewer pixels,lower requirements for computing power,low power consumption and low image acquisition cost,and is more suitable for the acquisition of star spot images.This topic explores the performance of compressed sensing for laser-based laser communication systems and the performance of atmospheric turbulence for spot images.The main research contents are as follows:1.The physical performance of compressed sensing applied to image acquisition of laser communication spots is studied.Investigated the different physical implementation forms of compressed sensing,designed the laser communication system spot image compression acquisition system,analyzed the advantages and disadvantages of different implementation forms and the performance of the satelliteto-ground laser communication system spot image acquisition;2.Research on compressed sensing image reconstruction algorithm.Investigated the current image reconstruction algorithms.In order to improve the edge definition and texture details of the reconstruction algorithm,a wavelet transform-based convolutional neural network reconstruction algorithm was proposed to integrate the loss function of the original image and its wavelet secondary image.Training the network model further improves the image reconstruction effect,so that the compressed sensing reconstructed image can recover more texture and edge details;3.The influence of the information loss of the compressed sensing image compression process on the analysis of atmospheric turbulence is studied,the process of beam propagation in the atmosphere is analyzed and the process is simulated,and atmospheric parameters are controlled to simulate the beam degraded by atmospheric turbulence and then compressed and collected.This quantifies the impact of different reconstruction algorithms on the key parameters of turbulence analysis after the image is reconstructed;4.The method of using the compressed information to directly calculate the spot position offset is studied.In order to bypass the complex mathematical relationship between the observation information and the spatial image and directly solve the centroid offset of the spot,this paper proposes a method to solve the position information of the spot image by constructing a measurement matrix,so that the laser communication system can be used in a small Calculate the center of mass of the light spot directly in the context of noise.For the case of high noise,this paper designs a compressed sensing noise suppression centroid solution(CSD-Center Net)model based on deep learning based on the above matrix.The model can also suppress the influence of noise on the accuracy of the spot centroid calculation by the measurement matrix method under heavy noise.This subject is a preliminary application research on the application of compressed sensing to collect wireless laser communication spot images.After analysis,the physical implementation of the proposed compressed sensing can meet the requirements of laser communication to varying degrees;the information loss caused by compressed sensing will affect the atmospheric turbulence The various researches of the company have caused different degrees of impact,but the impact can be controlled within the allowable range through the reasonable selection of the sampling rate and the parameters of the laser communication system;through the method of combining the sampling matrix and CSD-Center Net proposed in this paper,it can be It basically meets the accuracy and real-time requirements of laser communication beacon light tracking;in addition,the multiple sparsity deep learning reconstruction algorithm based on wavelet decomposition proposed in this paper has leading image recovery performance,while also dealing with edge and texture details.It has a better reconstruction effect and greatly reduces the impact of compressed sensing on the picture. |
PDF Full Text Request