ZTE Communications ›› 2022, Vol. 20 ›› Issue (S1): 7-15.DOI: 10.12142/ZTECOM.2022S1002

• Research Paper • Previous Articles     Next Articles

Auxiliary Fault Location on Commercial Equipment Based on Supervised Machine Learning

ZHAO Zipiao1, ZHAO Yongli1(), YAN Boyuan1, WANG Dajiang2   

  1. 1.State Key Laboratory of Information Photonics and Optical Communictions, Beijing University of Posts and Telecomunications, Beijing 100876, China
    2.ZTE Corporation, Shenzhen 518057, China
  • Received:2020-04-01 Online:2022-01-25 Published:2022-03-01
  • About author:ZHAO Zipiao received the B.S. degree in communication engineering from Hebei Univiversity of Engineering, China in 2017. She is currently a Ph.D. candidate of Beijing Univiversity of Posts and Telecomunications (BUPT), China. Her research focuses on software defined optical networks and machine learning in optiacal networks.|ZHAO Yongli (yonglizhao@bupt.edu.cn) received the B.S. degree in communication engineering and Ph.D. degree in electromagnetic field and microwave technology from Beijing Univiversity of Posts and Telecomunications (BUPT), China in 2005 and 2010, respectively. He is currently a professor of the Institute of Information Photonics and Optical Communications, BUPT. From Jan. 2016 to Jan. 2017, he was a visiting associate professor with UC Davis, USA. He has published more than 300 papers in international journals and conference. Since 2015, he has become a senior member of IEEE. His research focuses on software defined optical networking, elastic optical networks, datacenter networking, and optical network security.|YAN Boyuan received the B.S. degree in communication engineering from Beijing Univiversity of Posts and Telecomunications (BUPT), China in 2015. He is currently pursuing his Ph.D. degree with BUPT. His research interests include software defined optical networking, service function chaining, and network resource allocation with machine learning.|WANG Dajiang received B.S. and M.S. degrees in mechanical and electronic engineering from East China University of Science and Technology , China, and Shanghai University, China in 1996 and 2001 respectively. By far, he has been working with ZTE Corporation for more than 17 years and persistently doing the job related to OTN intelligent management and control. Now he is the Director of Intelligent Optical Network Product Planning in ZTE and mainly responsible for researching autonomous network architecture and solution on OTN. In total, as the chief author, he has issued more than 60 patents in the fields of WASON, SDON, and AN, most of which had been authorized by CNIPA and USPTO and commercialized in ZTE OTN products. Meanwhile, cooperating with other experts in the field, he has also published more than 20 papers in international influential journals and conferences.
  • Supported by:
    National Natural Science Foundation of China (NSFC)(61822105);ZTE Industry?Academia?Research Cooperation Funds(2018110016001047)

Abstract:

As the fundamental infrastructure of the Internet, the optical network carries a great amount of Internet traffic. There would be great financial losses if some faults happen. Therefore, fault location is very important for the operation and maintenance in optical networks. Due to complex relationships among each network element in topology level, each board in network element level, and each component in board level, the concrete fault location is hard for traditional method. In recent years, machine learning, especially deep learning, has been applied to many complex problems, because machine learning can find potential non-linear mapping from some inputs to the output. In this paper, we introduce supervised machine learning to propose a complete process for fault location. Firstly, we use data preprocessing, data annotation, and data augmentation in order to process original collected data to build a high-quality dataset. Then, two machine learning algorithms (convolutional neural networks and deep neural networks) are applied on the dataset. The evaluation on commercial optical networks shows that this process helps improve the quality of dataset, and two algorithms perform well on fault location.

Key words: optical network, fault location, supervised machine learning