| Compared with traditional microwave technology communication,space optic al communication has the characteristics of better confidentiality,higher transmiss ion code rate,and richer spectrum resources.It is the development direction of realizing the next generation of ultra-high-speed and large-capacity communicatio n.Space optical communication light source is a laser,which has strong directivi ty.Only when the transceivers of the two communication parties are precisely ali gned and tracked stably,the information carried by the laser can be received.Th e optical axis detection and alignment of space optical communication is the key technology for stable communication.In coherent optical communication system,there are strict requirements for spatial matching,polarization matching and intensity distribution matching between the received signal beam and local oscillator beam.If the system fails to achieve reasonable matching with high performance,the receiving sensitivity will be seriously reduced or deteriorated.The optical axis alignment between signal light and local oscillator light is the basic condition to realize the high performance matching.This thesis focuses on how to detect the alignment degree of optical axis between received signal light and local oscillator light in coherent optical communication system.1.Compare and analyze the advantages and disadvantages of optical axis detection technology based on Charge Coupled Device(CCD),Position Sensitive Detector(PSD),and Four Quadrant Detector(4QD),and the optical axis detection design scheme is based on 90°spatial optical mixer spatial light splitting in coherent optical communication is given.The optical axis detection process of the design scheme is theoretically deduced to prove the feasibility and effectiveness of the optical axis detection.2.There are various random noises contained in the received electrical signal.When the signal light is very weak,these noises will seriously affect the accuracy of the optical axis deviation detection.In response to this situation,analyze the types and sources of noise in detail,and model various types of noise,and a de-noising scheme using Kalman filter to suppress the noise of electrical signals is proposed.3.Analyze the advantages and disadvantages of traditional optical axis deviation detection algorithms based on fitting and database.The calculation amount and calculation accuracy of these algorithms are difficult to meet the optimal at the same time,and the inhomogeneity of the optical signal that seriously affects the detection accuracy cannot be suppressed and compensated.In the circumstances,propose an optical axis deviation detection algorithm based on BP neural network(Back Propagation Neural Network).4.Build the optical and electrical experimental platform according to the design scheme.The simulation and FPGA implementation of Kalman filter noise preprocessing show that the processed electrical signal noise has been effectively suppressed.According to the design,an experimental platform is built to collect data,and the designed neural network algorithm is used for training.Compared and analyze the effects of different training algorithms and transfer functions,and select the best applicable training algorithm and transfer function of the algorithm model.After training,we test the neural network,and the accuracy of optical axis deviation detection in the azimuth channel reaches 10-3mm,which verifies the performance of the designed optical axis deviation detection algorithm. |
PDF Full Text Request