Research On Signal Monitoring And Equalization Technologies In Optical OFDM System

With the vigorous development of internet services,optical access networks are facing needs and challenges of large capacity,high speed and flexible allocation of bandwidth.Orthogonal Frequency Division Multiplexing(OFDM),as a new type of optical modula?tion technology,is promising in the next-generation high-speed large-capacity optical ac-cess system due to its strong anti-dispersion,anti-fiber nonlinearity and high bandwidth utilization.Meanwhile,an adaptive optical OFDM communication system can flexibly al-locate bandwidth resources based on traffic requirements,which is a powerful technology to promote the development of the intelligent optical access network.The technologies of signal monitoring and equlatization are key to realize the flexible resource allocation and the improved performance in the adaptive optical OFDM communication system.Thus,the thesis focuses on the technologies of signal monitoring and equalization.The signal monitoring includes the identification of modulation format and the estimation of optical signal to noise ratio(OSNR).Firstly,the density peak-based clustering algorithm and the K-nearest neighbor regression(KNN)algorithm were proposed to find the cluster centers of the received signal's constellation and adaptively calculate the number of cluster centers as the standard of the identification.This method does not require training se-quences and has the advantage of blind identification.Secondly,the method of monitoring the OSNR based on the high-order moment with high precision was proposed.The OSNR of the multi-ary quadrature amplitude modulation(M-QAM)signal was calculated through extracting the points on the inner amplitude circle of the constellation,which solves the issue that the coventional high-order moment method is only accurate to estimate the OSNR of the M-QAM signal in a small OSNR range.For signal equalization,the density peak-based clustering algorithm and the K-means clustering algorithm were proposed to find the cluster centers of the received signal's constellation and make the optimum deci-sion for symbols.The proposed technology of signal equalization does not depend on the modulation format,where the system's bit error rate performance can be improved.