Performance Analysis And Optimization Of Optical Fiber Communication System In The Mlse Equalizer

Maximum likelihood sequence estimation is a technology based on Viterbi algorithm, it is used to improve the performance of ISI channel. For the additive white Gaussian noise channel, MLSE is considered to have the best performance.For high-speed optical fiber communication systems, dispersion is the main constraint factor, which limits the system's transmission capacity. Dispersion effects will lead to optical pulse stretcher, which can cause ISI, resulting in the closure of the received signal eye diagram, increasing the system bit error rate, reducing the transmission distance, reducing the performance of optical fiber communication system. In addition to traditional dispersion compensation fiber, the telecommunications channel equalization techniques can be used to achieve dynamic dispersion compensation, MLSE is one of them. With the signal processing technology and the inherent cost advantages of integrated circuits, the major fiber optic equipment manufacturers have the technology to carry out a large number of studies. The research of MLSE equalization in optical fiber communication system starting in 2005 gradually increased, both in theory and simulation of system performance studies, including experimental verification.This article focuses on performance analysis and optimization of MLSE, the main work of this article is:First introduced the digital fiber optic communication systems, optical transmitter and optical receiver of the main structure, analysis of various components of the system played a role and bring negative impact. Then introduced Viterbi algorithm decoding process and analyzed the decision sequence estimation formula with the additive white Gaussian noise channel.Analyzed a digital fiber optic communication system model, the noise of the model comes mainly from the thermal noise from the receiver. Analyzed the performance of MLSE with simulation in this model. Formula in the theory, propose a algorithm with better performance. However, the algorithm is more complicated, for the purpose of simplifying the algorithm, this paper proposed two optimization methods, and found a best performance of the optimization method by comparing.Finally, Analyzed a system model of which the noise is mainly the amplified spontaneous emission noise. At the receiver, the signal is not Gaussian, but CHi square distribution which is more complex in this model. Analyzed the noise and dispersion on system performance in this model. Using the previous model optimized method, studied the performance of these optimized methods with simulation. Finally, proposed two algorithm to simplified the decision formula, and compare the performance of these two methods.