基于神经网络的宽带线性调频信号幅相均衡  

Channel Equalization for Wide-band Linear Frequency Modulation Signal Based on Neural Network

作　　者：王克非 舒汀[1] 郁文贤[1] WANG Kefei;SHU Ting;YU Wenxian(Shanghai Key Laboratory of the Intelligent Sensing and Recognition,Shanghai Jiao Tong University,Shanghai 200240,China)

机构地区：[1]上海交通大学上海市智能探测与识别重点实验室

出　　处：《现代雷达》2019年第12期30-34,39,共6页Modern Radar

基　　金：国家自然科学基金资助项目(61571294);航空科学基金资助项目(2015ZD07006)

摘　　要：通道失衡是指多通道信号系统中,信号经过不同信道时由于各个信道的特性差异导致多通道信号产生不同失真的现象。为了对这种失真现象进行补偿和均衡,文中针对宽带雷达中常用的线性调频信号研究了一种基于神经网络的通道幅相一致性补偿算法,在均衡性能方面与传统算法进行了比较,并研究该算法在工程实现上的途径和技术难点。研究结果表明,该算法理论上有着优于传统算法的性能,并有较好的工程应用前景。Channel distortion refers to the distortion as a result of different features of different channels in multi-channel system.To compensate and equalize this distortion,this paper study on neural network algorithm in equalization of linear frequency modulation signal which is widely applied in wide-band radar,and compare with traditional method in terms of equalization performance.The method and difficult points in engineering implementation is also discussed.The research result shows that the algorithm perform better than existing method.Also there is a considerable application prospect in engineering.

关 键 词：频域均衡 幅相一致性 线性调频 神经网络 

分 类 号：TN957.51[电子电信—信号与信息处理]