面向智慧交通的神经网络辅助OFDM-IM信道估计方法  

A DNN Based OFDM-IM Channel Estimation Scheme for Intelligent Transportation Systems

作　　者：张思宇[1,2] 董青尚 张月霞[1,3] 赵阳[1] ZHANG Siyu;DONG Qingshang;ZHANG Yuexia;ZHAO Yang(School of Information and Communication Engineering,Beijing Information Science and Technology University,Beijing 102206,China;Key Laboratory of Information and Communication Systems,Ministry of Information Industry,Beijing Information Science and Technology University,Beijing 102206,China;Key Laboratory of Modern Measurement and Control Technology,Ministry of Education,Beijing Information Science and Technology University,Beijing 102206,China)

机构地区：[1]北京信息科技大学信息与通信工程学院,北京102206 [2]北京信息科技大学信息与通信系统信息产业部重点实验室,北京102206 [3]北京信息科技大学现代测控技术教育部重点实验室,北京102206

出　　处：《移动通信》2024年第10期78-85,共8页Mobile Communications

基　　金：国家自然科学基金青年项目“面向高铁通信的极化码编码索引调制传收理论与方法研究”(62301058);北京市教育委员会科技计划一般项目“面向6G大规模物联网的低功耗多维索引调制理论研究”(KM202411232010)。

摘　　要：作为新一代通信技术,正交频分复用索引调制(OFDM-IM)具有良好的抗子载波间干扰能力与较高的频谱能量效率,在6G协作智慧交通系统(C-ITS)中应用前景广阔。但OFDM-IM由于其载波排布结构,对高速衰落信道的估计性能较差,影响ITS通信可靠性。针对此问题,提出了一种基于索引导频(IP)和神经网络的OFDM-IM信道估计方法。所提方法通过基于IP的时域可靠测试频域插值算法实现信道的精确初始化,并通过深度神经网络(DNN)弥补初始估计中的非线性缺陷与插值误差。仿真结果显示,在IEEE 802.11p帧结构下,所提算法信道估计精度较现有非深度学习辅助方法在不同交通场景与通信参数下均有较为明显的提高。As a new generation of communication technology,orthogonal frequency division multiplex with index modulation(OFDM-IM)has good anti-inter-carrier-interference capability and high spectral efficiency,which has broad prospects in 6G cooperative intelligent transportation systems(C-ITS).Nevertheless,due to its carrier arrangement structure,OFDM-IM performs poorly in estimating high-speed fading channels,affecting the reliability of C-ITS.To address this problem,an OFDM-IM channel estimation method is proposed based on index pilots(IP)and neural networks.The proposed scheme achieves accurate channel initialization through time-domain reliable test and frequency-domain interpolation based on IP,and utilizes deep neural networks to compensate the non-linearity and interpolation errors in the initial estimation.Simulation results show that based on IEEE 802.11p framework,the channel estimation accuracy of the proposed scheme is significantly improved under different traffic scenarios and communication parameters compared with the non-deep learning aided methods.

关 键 词：智慧交通系统 正交频分复用 信道估计 索引调制 深度学习 

分 类 号：TN929.5[电子电信—通信与信息系统]