Study On Non-uniform Quantization Scheme In IMDD-UFMC Optical Transmission System

Recently,with the advancement of the emerging applications and services,such as AI,Big Data,cloud computing,and other technologies,it has increased rapidly for the demand of access bandwidth's size and type.To some extent,future transmission network system is required to meet the requirements of diversified transmission,including high-capacity,asynchronous transmission,high spectrum efficiency and others.Optical fiber access network system,based on Intensity Modulation Direct Detection(IMDD)multi-carrier modulation technology,has aroused widespread attention in the industry,which has own unique advantages like low cost,good anti-multipath effect,flexible and separable bandwidth.Therefore,the system becomes one of the main technologies in the current research direction of optical fiber access network.Moreover,multi-carrier modulation technology is one of current research hotspots of 5G air interface technology.Among the available multicarrier solutions,the Universal Filtered Multi-Carrier(UFMC),as a promising scheme,has become one of the strong candidate waveforms in 5G multi-carrier transmission technology.UFMC weights the pros and cons between filter bank multi-carrier modulation and orthogonal frequency division multiplexing.In order to achieve large-capacity data access services in 5G,wireless fronthaul access that based on optical fiber links is proposed,and it has become a hot spot for current research on access network technology.In the 5G access bearer solution of the optical fiber network,the optical transmission system with the IMDD multi-carrier modulation has attracted wide attention for its simple structure,easy implementation,and the overall merits of multi-carrier technology.However,based on IMDD-UFMC,high Peak to Average Power Ratio(PAPR)in the optical transmission system improves the high-resolution requirements of digital-to-analog converter/analog-to-analog converter(DAC/ADC).The high-resolution DAC/ADC caused by the high cost and high-power consumption will increase the system cost and complexity to a certain extent.Thus,it becomes a research hotspot about how to reduce quantization resolutions while maintaining system performance or even improving performance in this field.In order to effectively solve above problems,this thesis mainly focused on the study about the quantization technical solutions of the DAC in the IMDD-UFMC optical transmission system.The main research contents are as follows:1.This thesis introduces the research background of quantization schemes in IMDD-UFMC optical transmission system,and then analyzes research status,advantages,and disadvantages of quantization schemes in the multi-carrier system.After that,this thesis proposes that the research about a non-uniform quantization scheme which improves transmission performance and reduces system cost is an urgent problem to be solved in the IMDD-UFMC optical transmission system.Lastly,main contents are summarized.2.To solve the high PAPR problem in IMDD-UFMC system,a new quantitative model of joint clipping and quantization was proposed,and an IMDD-UFMC transmission system simulation platform was established to verify the feasibility of this model.The results show that this kind of quantization scheme reduces the PAPR of the UFMC system.In addition,the result of the simulation will be applied to the evaluation standard of the subsequent non-uniform quantization scheme.3.Based on analyzing the distribution characteristics of the UFMC signal,there are three nonuniform quantization schemes of signal distribution proposed: non-uniform quantization scheme based on Gaussian distribution,non-uniform quantization scheme based on nonparametric histogram estimation(NPHE)and non-uniform quantization scheme based on kernel density estimation(KDE).The feasibility of these three schemes is verified in the IMDD-UFMC system simulation platform of different optical fiber transmission distances.These schemes achieve significant performance improvement.The results show that all of these can improve the transmission performance of the IMDD-UFMC system in varying degrees.The order of improvement effect: KDE >NPHE >Gaussian distribution.4.To solve the problems that the computational complexity of the non-uniform quantization scheme based on signal estimation increases sharply with the increase of quantization resolution and the difficult estimation of solution time,three non-uniform quantization schemes based on clustering algorithm are proposed to the IMDD-UFMC system: the non-uniform quantization scheme based on the improved K-means clustering,non-uniform quantization scheme based on Gaussian Mixture Model(GMM)and non-uniform quantization scheme based on Self-Organization Mapping(SOM)).These clustering-based quantization schemes quantify signals without clipping process,which reduce the complexity of the system and control the solution time.The feasibility of these schemes is verified on the IMDD-UFMC-based optical transmission system simulation experiment platform with a transmission rate of 10 GB/s.The results of the simulations and experiments show that these schemes based on clustering algorithm are capable to reduce at least one DAC quantization bit,and it can be guaranteed that the transmission performance of the system with the limitation of quantization bit.This thesis proposed various non-uniform quantization schemes can effectively improve the quantization performance of the low quantization bit DAC.What's more,the problems of high-power consumption and high-cost caused by the use of high resolution DACs can be solved in different degrees.Thereby,these schemes implemented effectively ensure the performance of the IMDDUFMC optical transmission system in the process of limited quantization bits.