DCO-OFDM系统中BPNN信道估计方法  

BPNN Channel Estimation Method for DCO-OFDM System

作　　者：李晓记[1,2] 高天 阎威龙 杜卫海 LI Xiaoji;GAO Tian;YAN Weilong;DU Weihai(Key Laboratory of Cognitive Radio and Information Processing of Ministry of Education,Guilin 541004,Guangxi,China;School of Information and Communication,Guilin University of Electronic Technology,Guilin 541004,Guangxi,China)

机构地区：[1]桂林电子科技大学认知无线电与信息处理教育部重点实验室,广西桂林541004 [2]桂林电子科技大学信息与通信学院,广西桂林541004

出　　处：《实验室研究与探索》2022年第3期54-58,共5页Research and Exploration In Laboratory

基　　金：教育部第二批“新工科”研究与实践项目(E-XTYR20200653);教育部高等学校电子信息类教学指导委员会2020年度教学改革研究重点项目(2020-ZD-02);教育部产学合作协同育人项目(201902180021);广西高等教育本科教学改革工程项目(2021JGZ121)。

摘　　要：直流偏置光正交频分复用(DCO-OFDM)系统中的削波操作、LED的非线性特性以及光信号受水下阻挡物引起的多径效应,都会造成系统性能的下降。将后向传播神经网络(BPNN)应用于水下可见光通信系统以补偿由上述三方面因素引起的非线性失真,即先用最小二乘算法进行初步估计,再将估计值输入神经网络训练进一步提高估计精度,待训练完成后将所得模型运用于在线估计。通过仿真以及水下可见光通信系统测试平台分别测试不同路径数以及传输距离对算法性能的影响。实验结果表明,BPNN算法可取得比LS算法更好的估计精度。The clipping operation in direct current optical orthogonal frequency division multiplexing systems,the nonlinearity of the LED,and the multipath effect of the light signal caused by the underwater barrier can degrade the system performance.In this paper,backpropagation neural network is applied to underwater visible light communication system to compensate the nonlinear distortion caused by the above three aspects.Firstly,the least squares algorithm is used to make preliminary estimation,and then the estimated values are input into neural network to further improve the estimation accuracy.After the training is completed,the model is applied to online estimation.The influence of path number and different transmission distance on the algorithm is tested by simulation and underwater visible light communication system test platform.Experimental results show that the BPNN algorithm can achieve better estimation accuracy than the LS algorithm.

关 键 词：水下可见光通信 直流偏置光正交频分复用 非线性效应 信道估计 

分 类 号：TN929.12[电子电信—通信与信息系统]