Abstract

Optical performance monitoring technology (OPM) is a technology to monitor the physical state and transmission quality of optical network. In this paper, a scheme based on asynchronous delay tap histogram (ADTP) is proposed. On this basis, Convolutional Neural Network (CNN) and Support Vector Machine (SVM) are used to monitor optical performance, which completed modulation format recognition (MFI) and network performance parameter estimation under different damage. In this paper, an optical signal simulation system for 112 Gbps of DP-QPSK, DP-16PSK and DP-16QAM is set up. The optical signal-to-noise ratio (OSNR) and chromatic dispersion (CD) are mainly monitored. Asynchronous delay tap histogram is formed by asynchronous delay sampling. After extracting the characteristic parameters by CNN, the modulation format recognition, parameter OSNR and CD estimation are made. The simulation results show that CNN performs well in modulation format recognition, OSNR and CD estimation, with 100% modulation format recognition accuracy, 95% OSNR estimation accuracy, 95% CD estimation accuracy, 3.44% higher than asynchronous amplitude histogram (AAH)+SVM OPM scheme, and 40.79% higher than ADTP+CNN OPM scheme.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call