面向低冗余弹性光网络的两阶段多参量监测方法数据集

提出二阶段-多任务神经网络方法提升低冗余弹性光网络中光性能监测的准确率与可靠性。在第一阶段，通过将实验数据进行IQ补偿、色散补偿、CMA均衡后获得的信号幅度柱状图作为多任务神经网络的输入获得监测所得的OSNR和调制格式信息；在第二阶段，通过设定阈值将监测结果分为可靠结果与可疑结果并进行相应处理。利用可调谐激光器、外部IQ调制器、相干光电探测器、数字信号示波器进行数据采集,采用Matlab、Python软件对采集数据进行处理、通过Origin软件做图获得论文中的相关实验结果图片。

A two-stage multi-task neural network method is proposed to improve the accuracy and reliability of optical performance monitoring in low-redundancy elastic optical networks. In the first stage, the signal amplitude histogram obtained by performing IQ compensation, dispersion compensation and CMA equalization on experimental data is used as the input of the multi-task neural network to acquire the monitored OSNR and modulation format information. In the second stage, the monitoring results are divided into reliable results and suspicious results by setting thresholds, and corresponding processing is conducted. Data collection is implemented using a tunable laser, external IQ modulator, coherent photodetector and digital signal oscilloscope. The collected data is processed with Matlab and Python software, and relevant experimental result images in the paper are generated via Origin software.