ID 原文 译文
52137 在现实网络中采集的数据往往有分布不均匀、数据缺失等问题，如何构建完备的数据集是基于人工智能技术实现电力骨干通信网中告警预测的一个关键问题。 The data collected in the real network often has problems such as uneven distribution and missing data. How to construct a complete data set is a key issue based on artificial intelligence technology.
52138 文章提出了一种针对电力骨干通信网中告警预测的数据预处理方法，采用数据增强技术完成了数据的类均衡，使得最终的数据集符合模型输入的数据尺寸。 This paper proposes a data preprocessing method for alarm prediction in power backbone communication network. Data enhancement technology is used to complete the class balance of data, so that the final data set conforms to the data size input by the model.
52139 仿真结果表明，基于数据增强的告警预测数据预处理方法能够在较大程度上提高告警预测的准确率。 The simulation results show that the data pre-processing method based on data enhancement can improve the accuracy of alarm prediction to a large extent.
52140 基于信号功率非线性变换，结合神经网络，文章提出了一种利用深度神经网络(DNN)实现光信噪比(OSNR)监测的方法。 This paper proposes and demonstrates an Optical Signal Noise Ratio(OSNR) monitoring scheme using the signal power nonlinear transformation and Deep Neural Networks(DNN).
52141 通过对信号功率、信号2次方、4次方和8次方运算后对应的快速傅里叶变换(FFT)后的幅度获取信号与OSNR相关特征量，并利用DNN提取相关特征量以实现OSNR监测。 The features of signal after 2th, 4th, and 8th transformation and corresponding Fast Fourier Transformation(FFT) depend on the OSNR of signal. By utilizing the DNN to extract those OSNR depended specific features, the OSNR value can be estimated.
52142 仿真结果表明，针对28 Gbaud偏振复用(PDM)正交相移键控(QPSK)、PDM-8相移键控(PSK)、PDM-8正交振幅调制(QAM)和PDM-16QAM信号相干光通信系统，对应的背靠背OSNR监测平均标准误差分别为0. 10、0. 09、0. 33和0. 46 dB。对这4种调制格式，在入纤功率分别为4、4、3和3 dBm，传输距离分别为2 000、1 040、1 040和800 km单模光纤时，获得的OSNR监测平均标准误差分别为0. 43、0. 34、0. 66和0. 79 dB。 Simulation results for 28 Gbaud Polarization Division Multiplexing(PDM)-Quadrature Phase Shift Keying(QPSK), PDM-8 Phase Shift Keying(PSK), PDM-8 Quadrature Amplitude Modulation(QAM) and PDM-16 QAM signals show that the OSNR monitoring with mean estimation standard errors of 0. 10, 0. 09, 0. 33 and 0. 46 dB in back-to-back case and 0. 43, 0. 34, 0. 66 and 0. 79 dB in 2 000, 1 040, 1 040 and 800 km single mode fiber transmission case with input optical power of 4, 4, 3 and 3 dBm, respectively.
52143 基于互补氧化物半导体(CMOS)相机卷帘快门效应的可见光通信系统比特速率越高，拍摄的图片中每个比特对应的像素列数越少，且像素灰度值波动加剧，严重影响数据解调性能。 In Complementary Metal Oxide Semiconductor(CMOS) camera-based visible light communication system using the rolling shutter effect, the number of pixel columns representing one bit will decrease and the fluctuation of grayscale values became serious with the increasing of system bit rate, which leads to the degradation of demodulation performance.
52144 针对这一问题，文章提出采用全局和局部两个动态阈值判决结合分段降采样的解调算法， A novel demodulation algorithm based on global and local dynamic thresholds and segmented down sampling is proposed.
52145 首先由图片中全部待解调列的灰度值拟合得到全局动态阈值，判决后检测出所有数据帧头位置， The global dynamic threshold is fitted by grayscale values of all columns in an image, and frame headers contained in the image are located.
52146 然后根据各帧对应列的灰度值分别拟合出局部阈值， Then the local dynamic thresholds are fitted respectively by the grayscale values in each data frame.