一种光网络通信软故障检测方法及装置数据集

本数据集中的发明专利利用集成相干接收机获取待测光网络的数字残差谱，并利用自动编码器提取特征与重构原输入谱，从而利用重构误差判断是否存在软故障，完成检测任务，同时实现检测模型的高泛化性。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Python软件对采集数据作图获得专利中的相关实验结果图片。数据量为：1.39MB。

The invention patent included in this dataset uses an integrated coherent receiver to acquire the digital residual spectrum of the optical network under test, and employs an autoencoder to extract features and reconstruct the original input spectrum. Thereby, it identifies the presence of soft faults via the reconstruction error to fulfill the detection task, while achieving high generalization performance of the detection model. Data acquisition is conducted using a tunable laser, external IQ modulator, coherent photodetector, and digital signal oscilloscope. Relevant experimental result images in the patent are generated by plotting the acquired data with Python software. The size of this dataset is 1.39 MB.

搜集汇总

背景与挑战

背景概述

该数据集涉及光网络通信软故障检测方法及装置，利用集成相干接收机和自动编码器技术进行故障检测，数据量为1.37MB，包含21个文件。数据集由北京邮电大学孙凯旋发布，属于电子与通信技术领域。