基于数字信号处理辅助相干检测的高频谱效率光纤传输技术

The capacity limit of traditional WDM is nearly reached with the growth of bandwidth requirement. Therefore, digital-signal-processing (DSP) aided optical coherent detection technology is very necessary to study for improving the spectral efficiency of fiber. However, the spectral efficiency of Nyquist-WDM and CO-OFDM is confined of the orthogonal condition. And there is a big gap between their maximum spectral efficiency and the Shannon limit. The purpose of the project is to thoroughly study the fiber media's Shannon limit, and increase the spectral efficiency of fiber for the information infrastructure by the means of advanced DSP technology. We have relized over 100 Gb/s optical coherent single-carrier and OFDM transmission. The project is plan to explore the nonlinear Shannon limit of high spectral efficiency fiber transmission in theory and by experiments,propose the faster-than-Nyquist transmission theory in the fiber and verify it by experiment, study the polarization demultiplexing technology of independent component analyze algorithms and novel high spectral efficent modulation formats, so as to explore the basic law of ultra dense WDM, overcome the damage to dense WDM channel, approach the capacity limit of fiber media. The result of the project will be of great significance to the ultra high capacty and high speed information transmission.

随着带宽需求快速增长，传统WDM传输方式容量极限已逐渐凸显，因此研究通过数字信号处理辅助的相干检测技术提高WDM系统频谱效率非常必要。但目前的Nyquist-WDM和CO-OFDM技术受正交性条件局限，能达到的频谱效率距香农极限还有很大差距。本课题目的是通过对光纤介质的香农极限进行系统的基础理论研究，在现有信息基础设施和技术条件上，以先进的DSP算法辅助提升光纤传输的频谱效率。本课题组已实现超过100Gb/s的单载波和OFDM相干光通信实验。本课题在此基础上从理论和实验探索高频谱效率光纤传输的香农极限，创新性地在光纤传输领域提出超Nyquist传输理论并进行实验验证，研究在偏振复用系统中采用基于ICA算法的解复用技术，研究新型的高频谱效率光信号调制格式，从而探索超密集WDM的基本规律，克服超密集WDM信道的损伤，逼近光纤介质的容量极限。本课题的研究成果将对解决大容量传输的难题具有重要意义。

在国家基金的支持下，围绕探索高频谱效率光纤传输的香农极限，研究高速光通信系统中的新型调制格式等一系列问题，本项目组按照申请书的计划开展研究工作，并取得了预期的研究成果，这些成果都通过实验验证，包括：探索高频谱效率光纤传输的非线性香农极限；超Nyquist信号的传输及均衡；基于独立成分分析的偏振解复用技术；以及新型的高频谱效率光信号调制格式的研究。本项目组共发表论文30篇，其中，SCI论文28篇，EI论文2篇，申请国家发明专利5项，已授权国家发明专利1项。培养硕士生25名，博士生10名。