基于数据驱动的光纤通信链路状态监测  

Data driven condition monitoring of optical fiber communication link

作　　者：余晓丽[1] 徐慧[1] YU Xiaoli;XU Hui(Nanchang Institute of Technology,Nanchang 330044,China)

机构地区：[1]南昌理工学院,南昌330044

出　　处：《激光杂志》2023年第2期159-163,共5页Laser Journal

基　　金：江西省教育厅科技项目(No.GJJ191016)。

摘　　要：为提升光纤通信可靠性,掌控光纤通信链路状态变化情况,提出了基于数据驱动的光纤通信链路状态监测方法。首先采集光纤通信链路状态信号,并对光纤通信链路状态信号进行噪声去除操作,然后利用小波算法提取链路状态特征,运用改进概率神经网络依据链路状态特征设计光纤通信链路状态监测模型,最后进行了链路状态监测仿真测试,结果表明:该方法可以有效去除光纤通信链路状态信号噪声,监测出各种光纤通信链路状态,光纤通信链路状态识别正确率超过95%,光纤通信链路状态偏差小于5%,识别时间控制在6 s以内,完全可以满足光纤通信的实际应用要求。In order to improve the reliability of optical fiber communication, control the state changes of optical fiber communication link, an optical fiber communication link condition monitoring method based on data-driven method is proposed. Firstly, the optical fiber communication link state signal is collected, and the noise of the optical fiber communication link state signal is removed. Then the link state characteristics are extracted by wavelet algorithm. the probabilistic neural network is improved, and the optical fiber communication link state monitoring model is designed according to the link state characteristics, Finally, the simulation test of link state monitoring is carried out. The results show that this method can effectively remove the signal noise of optical fiber communication link state and monitor various optical fiber communication link states. The accuracy of optical fiber communication link state identification is more than 95%, the deviation of optical fiber communication link state is less than 5%, and the identification time of optical fiber communication link state is controlled within 6s, It fully meets the practical application requirements of optical fiber communication.

关 键 词：数据驱动 光纤通信技术 链路状态监测 去噪处理 提取特征 

分 类 号：TN929[电子电信—通信与信息系统]