摘要
传统的正交频分复用技术对正交性要求高且存在频谱资源的浪费,不适合大规模物联网接入5G网的应用场景,偏移正交幅度调制滤波器组多载波技术是该应用场景的一个解决方案,但该技术应用于40~100Gbps光纤主干网时存在固有干扰和光信道噪声的问题,为此提出了基于小波神经网络的物联网主干光信道估计方法。通过最小二乘法对光信道进行初步估计,基于估计的结果和布谷鸟搜索算法初始化小波神经网络的参数;通过伪逆运算动态更新网络的权重参数,从而对信道状态进行实时跟踪,保持较高的估计准确性。实验结果显示,估计方法的效果好于传统的最小二乘估计和最小均方误差估计。
The traditional orthogonal frequency division multiplexing technique suffers from high demand of orthonormality and frequency resource waste,so that it is not suitable for the scenario of large scale Internet of things accessing 5 G network,filterbank based multiple carrier with offset quadrature amplitude modulation technique is a solution for this scenario,but inherent interference and channel noise come up when this technique is applied to 40~100 Gbps fiber backbone networks,in view of this issue,a backbone optical channel estimate method for internet of things is proposed,it is based on wavelet neural network.First of all,it takes advantage of least square method to estimate optical channel,and the estimated result and cuckoo search algorithm are used to init the parameters of wavelet neural network.Then,the pseudo inverse operation is adopted to dynamically update the weight parameters of the network,so that the channel states are real time tracked to keep high estimate accuracy.Experimental results indicate that the proposed estimate method outperforms the traditional least square estimate and minimum mean squared error estimate.
作者
李扬
LI Yang(Department of Electronics and Information,Jiangsu Vocational College of Business,Nantong 226000,China;Jiangsu Province Internet of Things and Visual Intelligent Processing Engineering Technology Research and Development Center,Nantong 226000,China)
出处
《光学技术》
CAS
CSCD
北大核心
2020年第3期368-374,共7页
Optical Technique
基金
江苏高校‘青蓝工程’资助(201706-202006)
江苏省“333工程”科研项目(BRA2018220)
南通市科技计划项目(GY12015037)
南通市港闸区科技项目(GZKJ2018JHK008)。
关键词
小波神经网络
信道估计
大规模物联网
主干网
滤波器组多载波
第五代移动通信
wavelet neural network
channel estimate
big scale internet of things
backbone networks
filterbank based multiple carrier
fifth generation mobile communication
