电力信息与通信技术2025，Vol.23Issue(8)：32-40,9.DOI:10.16543/j.2095-641x.electric.power.ict.2025.08.05
基于改进DFT与DnGRUNet的HPLC信道估计
HPLC Channel Estimation Based on Improved DFT and DnGRUNet
摘要
Abstract
High speed power line carrier(HPLC)based on orthogonal frequency division multiplexing(OFDM)has been widely used in distribution networks and smart homes.The HPLC channel has problems such as complex noise,multipath propagation,and variable impedance,which have adverse effects on signal detection.This paper investigates the channel estimation problem of HPLC systems.Firstly,the approximate power of sampling point noise within the cyclic prefix is defined,and a DFT algorithm based on the approximate power of sampling point noise(NPDFT)is proposed to modify the channel impulse response obtained by traditional DFT algorithms.Secondly,based on NPDFT,a lightweight deep neural gated network(DnGRUNet)is proposed,which combines deep neural networks(DNNs)and gated recurrent units(GRUs).DNNs are used to extract and refine multipath channel features,while GRUs are used to learn channel changes between OFDM symbols.The simulation results show that the channel estimation error of the NPDFT-DnGRUNet algorithm is much lower than that of the comparison algorithm,which can effectively reduce the signal detection error rate.关键词
高速电力线载波通信/信道估计/正交频分复用/门控循环单元/离散傅里叶变换Key words
HPLC/channel estimation/OFDM/GRU/DFT分类
信息技术与安全科学引用本文复制引用
张良,张宇,刘永春,田景秋..基于改进DFT与DnGRUNet的HPLC信道估计[J].电力信息与通信技术,2025,23(8):32-40,9.基金项目
国家电网有限公司总部管理科技项目资助"面向海量负荷控制交互的安全可靠多元多域通信技术研究"(5108-202218280A-2-384-XG). （5108-202218280A-2-384-XG）
