Experimental Demonstration of Optical Fiber Eavesdropping Detection Based on Deep Learning

This paper proposes an effective eavesdropping detection scheme based on deep learning. The feasibility is verified in the experiment with 300km optical transmission. Results show that detection accuracy of eavesdropping can be achieved over 95%.

[1]  Chao Lu,et al.  An Optical Communication's Perspective on Machine Learning and Its Applications , 2019, Journal of Lightwave Technology.

[2]  Habib Fathallah,et al.  Optical fiber tapping: Methods and precautions , 2011, 8th International Conference on High-capacity Optical Networks and Emerging Technologies.

[3]  K. Shaneman,et al.  Optical network security: technical analysis of fiber tapping mechanisms and methods for detection & prevention , 2004, IEEE MILCOM 2004. Military Communications Conference, 2004..

[4]  Hao Ma,et al.  Optical Fiber Communication Network Eavesdropping and Defensive Measures , 2018 .