首页|基于改进卷积神经网络的可见光通信系统信道估计方法

基于改进卷积神经网络的可见光通信系统信道估计方法

扫码查看
为了提升可见光通信系统的通信质量,开展了基于改进卷积神经网络的可见光通信系统信道估计方法研究.首先,建立可见光通信系统信道模型,获取信道增益、LED限制带宽、频域响应的模等参数.其次,从可见光通信系统导频分布与导频间隔两个维度,设计系统导频结构,准确得出系统发送端想要传输的信道信息.利用改进卷积神经网络,开展可见光通信系统信道估计,求解信道估计结果.实验结果表明,该方法的信道估计误码率不存在大幅度波动,最高不超过0.5%,信道传输效果较好.
Channel estimation method of visible optical communication system based on improved convolutional neural network
In order to improve the communication quality of visible optical communication sys-tem,reduce the number of error codes generated by signal damage,improve the accuracy of channel estimation,optimize the channel transmission effect of the system channel,the im-proved convolutional neural network is introduced,and the channel estimation method of visi-ble optical communication system based on improved convolutional neural network is studied.First,the channel model of visible optical communication system is established to obtain param-eters such as channel gain,LED limit bandwidth and mode of frequency domain response.Sec-ondly,the pilot structure of the pilot distribution and the pilot interval of the visible optical communication system is designed to accurately obtain the channel information of the transmit-ting end of the system.On this basis,the improved convolutional neural network is used to de-velop the channel estimation of the visible light communication system and solve the channel es-timation results.The experimental analysis results show that after the proposed channel estima-tion method,the channel estimation error rate does not fluctuate greatly,the highest does not exceed 0.5%,the channel transmission effect is good,the estimation accuracy and the commu-nication quality level of visible optical communication system are significantly improved.

improving convolutional neural networkestimationvisible optical communica-tion systemmethodchannelbit error rate

赵小强

展开 >

郑州工业应用技术学院信息工程学院,河南新郑 451100

改进卷积神经网络 可见光通信系统 方法 信道 误码率

河南省科学技术厅(省级)

172102210169

2024

长江信息通信
湖北通信服务公司

长江信息通信

影响因子:0.338
ISSN:2096-9759
年,卷(期):2024.37(2)
  • 5