高级检索

  • ISSN 1006-3080
  • CN 31-1691/TQ

基于实值神经网络的FBMC/OQAM系统PAPR降低方法

何超逸 袁伟娜

何超逸, 袁伟娜. 基于实值神经网络的FBMC/OQAM系统PAPR降低方法[J]. 华东理工大学学报(自然科学版). doi: 10.14135/j.cnki.1006-3080.20210621001
引用本文: 何超逸, 袁伟娜. 基于实值神经网络的FBMC/OQAM系统PAPR降低方法[J]. 华东理工大学学报(自然科学版). doi: 10.14135/j.cnki.1006-3080.20210621001
HE Chaoyi, YUAN Weina. PAPR Reduction Method of FBMC/OQAM System Based on Real Valued Neural Network PAPR Reduction Method of FBMC/OQAM System Based on Real Valued Neural Network[J]. Journal of East China University of Science and Technology. doi: 10.14135/j.cnki.1006-3080.20210621001
Citation: HE Chaoyi, YUAN Weina. PAPR Reduction Method of FBMC/OQAM System Based on Real Valued Neural Network PAPR Reduction Method of FBMC/OQAM System Based on Real Valued Neural Network[J]. Journal of East China University of Science and Technology. doi: 10.14135/j.cnki.1006-3080.20210621001

基于实值神经网络的FBMC/OQAM系统PAPR降低方法

doi: 10.14135/j.cnki.1006-3080.20210621001
基金项目: 国家自然科学基金(61501187)
详细信息
    作者简介:

    何超逸(1997-),男,陕西人,硕士生,主要研究方向为多载波通信系统的PAPR降低方法。E-mail:18229092069@163.com

    通讯作者:

    袁伟娜,E-mail:wnyuan_ice@163.com

  • 中图分类号: TN929.5

PAPR Reduction Method of FBMC/OQAM System Based on Real Valued Neural Network PAPR Reduction Method of FBMC/OQAM System Based on Real Valued Neural Network

  • 摘要: 偏移正交幅度调制滤波器组多载波(Filter Bank Multicarrier with Offset Quadrature Amplitude Modulation, FBMC/OQAM)系统是5G多载波通信系统候选方案之一,与正交频分复用(Orthogonal Frequency Division Multiplexing, OFDM)等多载波方案一样,存在峰均功率比(Peak-to-Average Power Ratio, PAPR)较高的问题,影响高功率放大器(High-Power Amplifier, HPA)的效率。针对FBMC/OQAM系统PAPR过高的问题,提出了一种基于实值神经网络的方法。该方法在发送端和接收端搭建两个实值神经网络,分别用来降低PAPR和误码率(Bit Error Ratio, BER)。仿真结果表明,相较于色散选择性映射(Dispersive Selected Mapping, DSLM)、限幅(Clipping)、编码以及PRnet方法,本文方法对PAPR和BER性能都有一定的提升。

     

  • 图  1  FBMC/OQAM系统发送端框图

    Figure  1.  Transmitter block diagram of FBMC/OQAM system

    图  2  FBMC/OQAM信号结构

    Figure  2.  FBMC/OQAM signal structure

    图  3  基于实值神经网络降低PAPR的系统结构

    Figure  3.  System structure of PAPR reduction based on real valued neural network

    图  4  PAPR降低模块结构图

    Figure  4.  Structure diagram of PAPR reduction module

    图  5  解压缩模块结构图

    Figure  5.  Structure diagram of decompression module

    图  6  不同PAPR降低方法的CCDF性能比较

    Figure  6.  CCDF performance comparison of with different PAPR reduction methods

    图  7  不同PAPR降低方法的BER性能比较

    Figure  7.  BER performance comparison of with different PAPR reduction methods

    图  8  不同$ {{\rm{PAPR}}}_{{\rm{pre}}} $的BER性能比较

    Figure  8.  BER performance comparison of different $ {{\rm{PAPR}}}_{{\rm{pre}}} $

    表  1  模型参数设置

    Table  1.   Parameters for models

    ParameterValues
    Num. of input neurons2
    Num. of FC layer1 neurons10
    FC layer1 activationtanh
    Num. of FC layer2 neurons5
    FC layer1 activationtanh
    Learning rate0.001
    Maximum number of iterations50000
    $ {\alpha }_{1},{\alpha }_{2} $0.001
    下载: 导出CSV

    表  2  模型复杂度比较

    ModelNumber of modelsModel structureReal mult.Real adds
    PRnet2[256 256 256 256 256]
    [256 256 256 256 256]
    73400327340032
    This paper2[10 5]
    [10 5]
    327680327680
    下载: 导出CSV
  • [1] ZAYANI R, SHAÏEK H, ROVIRAS D, et al. Closed-form BER expression for (QAM or OQAM)-based OFDM system with HPA nonlinearity over Rayleigh fading channel[J]. IEEE Wireless Communications Letters, 2015, 4(1): 38-41. doi: 10.1109/LWC.2014.2365023
    [2] ZAYANI R, SHAÏEK H, ROVIRAS D, et al. Lattice-reduction-aided equalization for MIMO-FBMC systems[J]. IEEE Wireless Communications Letters, 2019, 8(1): 101-104. doi: 10.1109/LWC.2018.2859942
    [3] WANG Y C, LUO Z Q. Optimized iterative clipping and filtering for PAPR reduction of OFDM signals[J]. IEEE Transactions on Communications, 2011, 59(1): 33-37. doi: 10.1109/TCOMM.2010.102910.090040
    [4] CHEN J. Low-PAPR precoding design for massive multiuser MIMO systems via Riemannian manifold optimization[J]. IEEE Communications Letters, 2017, 21(4): 945-948. doi: 10.1109/LCOMM.2017.2650218
    [5] 翁童, 袁伟娜. 一种基于SPSO算法降低FBMC系统PAPR的新方法[J]. 华东理工大学学报(自然科学版), 2020, 46(5): 702-708.
    [6] CHENG X, LIU D, SHI W, et al. A novel conversion vector-based low-complexity SLM scheme for PAPR reduction in FBMC/OQAM systems[J]. IEEE Transactions on Broadcasting, 2020, 66(3): 656-666. doi: 10.1109/TBC.2020.2977548
    [7] 伏威, 袁伟娜. 一种基于PTS方法降低FBMC系统PAPR的新方法[J]. 华东理工大学学报(自然科学版), 2019, 45(4): 639-645.
    [8] LI Z, JIN N, WANG X, et al. Extreme learning machine-based tone reservation scheme for OFDM systems[J]. IEEE Wireless Communications Letters, 2021, 10(1): 30-33. doi: 10.1109/LWC.2020.3019792
    [9] CHEN J, WEN C. PAPR reduction of OFDM signals using cross-entropy-based tone injection schemes[J]. IEEE Signal Processing Letters, 2010, 17(8): 727-730. doi: 10.1109/LSP.2010.2051617
    [10] KAGEYAMA T, MUTA O. A partial scrambling overlapped selected mapping PAPR reduction scheme for OFDM/OQAM systems[C]//2019 7th International Japan-Africa Conference on Electronics, Communications, and Computations, (JAC-ECC). Egypt : IEEE, 2019: 37-40.
    [11] K S S, Bulusu C, Shaiek H, et al. Reduction of PAPR for FBMC/OQAM systems using dispersive SLM technique[C]//2014 11th International Symposium on Wireless Communications Systems (ISWCS). Spain : IEEE, 2014: 568-572.
    [12] SOHN I, KIM S C. Neural network based simplified clipping and filtering technique for PAPR reduction of OFDM signals[J]. IEEE Communications Letters, 2015, 19(8): 1438-1441. doi: 10.1109/LCOMM.2015.2441065
    [13] KIM M, LEE W, CHO D H. A novel PAPR reduction scheme for OFDM system based on deep learning[J]. IEEE Communications Letters, 2018, 22(3): 510-513. doi: 10.1109/LCOMM.2017.2787646
    [14] LIU Z, HU X, HAN K, et al. Low-complexity PAPR reduction method for OFDM systems based on real-valued neural networks[J]. IEEE Wireless Communications Letters, 2020, 9(11): 1840-1844. doi: 10.1109/LWC.2020.3005656
    [15] BELLANGER M. Physical layer for future broadband radio systems[C]//2010 IEEE Radio and Wireless Symposium (RWS). USA: IEEE, 2010: 436-439.
  • 加载中
图(8) / 表(2)
计量
  • 文章访问数:  10
  • HTML全文浏览量:  6
  • PDF下载量:  7
  • 被引次数: 0
出版历程
  • 收稿日期:  2021-06-21
  • 网络出版日期:  2021-10-12

目录

    /

    返回文章
    返回