借助聚类分析的光发射机信号畸变评估方法 下载: 1106次
ing at the application of optical transmitter signal distortion analysis, we adopt a coherent detection system with slowly varying phase, analyze the preprocessing of the distorted optical signal by means of cluster analysis, and estimate the distortion coefficients accurately to achieve a better constellation restoration effect. In this system, we adjust the length of the local oscillator fiber link to be equal to the fiber length of the signal link, so that the phase difference between the local oscillator and the signal light can be approximately constant over a long time window. It means that we can do intelligent analysis and preprocessing of the constellation for the long time window data by means of cluster analysis. Then, we obtain the distortion coefficients and cluster centroids based on cluster analysis of signal constellation, and use the location information of cluster centroids to equalization process and carrier phase recovery. This scheme solves signal recovery problems in the presence of signal distortion and phase noise, enabling accurate signal distortion evaluation. In this system, we simulate and experiment three kinds of distorted signals (I-Q gain imbalance, I-Q phase error, and I-Q amplitude uneven distribution) by using cluster analysis and blind-phase-search (BPS) algorithm. The results show that in the experimental system we can accurately estimate the transmitter signal distortion mentioned by using cluster analysis.
周钟青, 杨彦甫, 张群, 何倩雯, 向前, 曹军涛. 借助聚类分析的光发射机信号畸变评估方法[J]. 激光与光电子学进展, 2018, 55(3): 030604. Zhongqing Zhou, Yanfu Yang, Qun Zhang, Qianwen He, Qian Xiang, Juntao Cao. Optical Transmitter Signal Distortion Evaluation Based on Cluster Analysis[J]. Laser & Optoelectronics Progress, 2018, 55(3): 030604.