| In the era of high-speed traffic development,the role of optical fiber communication technology in information transmission is growing.It provides strong support for the vast number of Internet users and services,and constantly improves people’s communication and life experience.However,in the context of such an information explosion,communication systems and networks are transmitting more and more information,and the demand for high speed and high capacity transmission systems is becoming more and more urgent.However,in the context of such an information explosion,communication systems and networks are passing more and more information,and the need for high speed and high capacity transmission systems is becoming more and more urgent.In this case,rates of hardware devices in the traditional signal electrical sampling mode requires higher and higher and this mode is increasingly difficult to implement in practical applications,which has gradually limited the system to detect high-speed signals.Therefore,all-optical sampling technology as a new direction of sampling technology has attracted a lot of attention from researchers.All-optical sampling technology achieves the sampling of the transmitted signal to be measured in the optical domain by means of short pulses with low repetition frequency,which solves the problem of traditional electrical sampling that requires high-speed balanced detectors and acquisition cards for signal conversion and reception.In this way,the technology greatly increases the upper limit of the signal transmission rate,and also greatly reduces the application cost of the system hardware equipment.Based on the full investigation and analysis of the development status and purpose of all-optical sampling technology,this thesis focuses on the linear optical sampling system,and introduces the hardware and software related modules of the system in detail.The golden section software synchronization algorithm,the single-step interpolation-based software synchronization algorithm and the third spline interpolation-based peak extraction algorithm are proposed respectively.The main research work of this thesis is as follows:(1)In order to improve the accuracy and reduce the computational complexity of software synchronization of sampled signals in linear optical sampling systems,a golden section software synchronization algorithm is proposed.The algorithm obtains the spectral information of the sampling point by fast Fourier transform and finds its coarse period value;and then performs spectral refinement by golden division in the local interval near the coarse period to obtain the exact number of periods of the sampling point at the desired spectral resolution,so as to recover a clear eye diagram of the signal.Experimental measurements are performed on polarization division multiplexed quadrature phase shift keying(PDM-QPSK)signals at five rates of 4 Gbps,8 Gbps,16 Gbps,32 Gbps,and 64 Gbps.Comparing the eye diagram and constellation diagram processed by the CZT algorithm,as well as their EVM values and Q-factors,it is demonstrated that the proposed algorithm performs better in recovering signal time domain information,with a maximum reduction of 0.84%in EVM value and a maximum improvement of 0.25 in Q-factor,as well as a 55%reduction in computational complexity.(2)In order to avoid constantly refining the spectrum to obtain the exact period and further reduce the computational complexity of software synchronization,a single step interpolation-based software synchronization algorithm is proposed.The algorithm finds the coarse period of the sampling points,uses the formula calculation to find the interpolation position near the coarse period,and directly calculates the compensation error value by one-step interpolation at both ends of the coarse period to obtain the exact period of the sampling points.The experimental measurements of PDM-QPSK signals at five rates of 4 Gbps,8 Gbps,16 Gbps,32 Gbps and 64 Gbps are compared with the results obtained by the golden section software synchronization algorithm and the CZT software algorithm,which also have better performance in recovering the signal time domain information.The EVM is reduced by 0.61%and the Q-factor is improved by 0.24.The computational complexity is reduced by 76%compared with the CZT algorithm,which greatly improves the efficiency of software synchronization.(3)In order to improve the accuracy of pulse peak points extraction in the linear optical sampling system and to ensure the performance of the subsequent software synchronization algorithm,a peak extraction algorithm based on cubic spline interpolation is proposed.By using three times spline interpolation for each pulse’s local interval of acquisition data processing to compensate for the limitation of low sampling frequency of the acquisition card,the exact peak points of each pulse is found.The peak extraction results before and after interpolation are compared through experiments on PDM-QPSK signals,as well as the eye diagram and constellation diagram of the signal reception results obtained by combining the subsequent software synchronization algorithm analysis.The experimental demonstrate the improved signal reception performance of the proposed algorithm,while its interpolation at local intervals of each cycle according to the pulse characteristics also greatly improves the computational speed of the algorithm. |
PDF Full Text Request