Wan Zhiquan, Yu Zhenming, Shu Liang, Zhao Yilun, Zhang Haojie, Xu Kun
Opt Express. 2019 Apr 15;27(8):11281-11291. doi: 10.1364/OE.27.011281.
An intelligent optical performance monitor using multi-task learning based artificial neural network (MTL-ANN) is designed for simultaneous OSNR monitoring and modulation format identification (MFI). Signals' amplitude histograms (AHs) after constant module algorithm are selected as the input features for MTL-ANN. The results obtained from simulation and experiment of NRZ-OOK, PAM4 and PAM8 signals demonstrate that MTL-ANN could achieve OSNR monitoring and MFI simultaneously with higher accuracy and stability compared with single-task learning based ANNs (STL-ANNs). The results show an MFI accuracy of 100% for the three modulation formats under consideration. Furthermore, OSNR monitoring with mean-square error (MSE) of 0.12 dB and accuracy of 100% is achieved while regarding it as regression problem and classification problem, respectively. In this intelligent optical performance monitor, only a single MTL-ANN is deployed, which enables reduced-complexity optical performance monitor (OPM) devices for multi-parameters estimation in future heterogeneous optical network.
设计了一种基于多任务学习的人工神经网络（MTL-ANN）的智能光性能监测器，用于同时进行光信噪比（OSNR）监测和调制格式识别（MFI）。选择经过常数模算法后的信号幅度直方图（AHs）作为MTL-ANN的输入特征。从非归零开关键控（NRZ-OOK）、四电平脉冲幅度调制（PAM4）和八电平脉冲幅度调制（PAM8）信号的仿真和实验结果表明，与基于单任务学习的人工神经网络（STL-ANNs）相比，MTL-ANN能够同时实现更高精度和稳定性的OSNR监测和MFI。结果显示，对于所考虑的三种调制格式，MFI准确率达到100%。此外，在分别将其视为回归问题和分类问题时，实现了均方误差（MSE）为0.12 dB且准确率为100%的OSNR监测。在这种智能光性能监测器中，仅部署了单个MTL-ANN，这使得未来在异构光网络中用于多参数估计的低复杂度光性能监测（OPM）设备成为可能。
