通信与信息网络学报

• Special Focus: UAV Communications • 上一篇    下一篇

复卷积神经网络设计及其在城市环境下无人机方向估计中的应用

  

  • 修回日期:2020-06-07 出版日期:2020-06-25 发布日期:2020-07-14

Complex-Valued Convolutional Neural Networks Design and Its Application on UAV DOA Estimation in Urban Environments

Bai Shi(),Xian Ma(),Wei Zhang(),Huaizong Shao(),Qingjiang Shi(),Jingran Lin()   

  1. University of Electronic Science and Technology of China, Chengdu 611731, China
    University of Electronic Science and Technology of China, Chengdu 611731, China;Science and Technology on Electronic Information Control Labaratory,Chengdu 610036,China
    University of Electronic Science and Technology of China, Chengdu 611731, China;Peng Cheng Laboratory,Shenzhen 518055,China
    Tongji University, Shanghai 200092, China
    University of Electronic Science and Technology of China, Chengdu 611731, China;Peng Cheng Laboratory,Shenzhen 518055,China
  • Revised:2020-06-07 Online:2020-06-25 Published:2020-07-14
  • About author:Bai Shi received his B.S. degree in communication engineering from Nanchang University in 2015 and his M.S.degree in electronic and communication engineering from University of Electronic Science and Technology of China (UESTC) in 2018. He is currently pursuing his Ph.D.degree in UESTC.|His research interests include communication signal processing.|Xian Ma received his B.S.degree in communication engineering from University of Electronic Science and Technology of China(UESTC)in 2019. Currently,he is pursuing his M.S.degree in UESTC.|His research interests include deep learning and communication signal processing.|Wei Zhang[corresponding author]received his B.S. degree in electronic engineering and his M.S.degree in circuits and systems from University of Electronic Science and Technology of China(UESTC),Chengdu China,in 2008 and 2011,respectively.Currently,he is pursuing his Ph.D.degree in information and communication engineering with the School of Information and Communication Engineering,UESTC.|His research interests include electromagnetic big data analysis and information intelligent processing.|Huaizong Shao received his M.S. degree in electrical engineering from Sichuan University, Chengdu, China, in 1998, and his Ph.D. degree in information and communication engineering from University of Electronic Science and Technology of China (UESTC),Chengdu, China, in 2003. Since 2003, he has been with the School of Communication and Information Engineering,UESTC,where he is currently a professor. From May 2014 to April 2015,he was a visiting scholar with the University of Sheffield,Sheffield,UK.|His research interests include communication and radar signal processing.|Qingjiang Shi received his Ph.D. degree in electronic engineering from Shanghai Jiao Tong University,Shanghai,China,in 2011. From September 2009 to September 2010, he visited Prof. Z.-Q. (Tom) Luo’s research group at University of Minnesota,Twin Cities. In 2011, he worked as a research scientist at Bell Labs China. From 2012,he was with the School of Information and Science Technology at Zhejiang Sci-Tech University. From February 2016 to March 2017,he worked as a research fellow at Iowa State University,USA.From 2018,he is currently a professor with the School of Software Engineering at Tongji University.|His interests lie in algorithm design and analysis with applications in machine learning,signal processing and wireless networks. So far he has published more than 40 IEEE journal papers (5 papers were ESI highly cited papers and one was nominated as the Best Paper Award of IEEE Signal Processing Society in 2016)and filed more than 20 national patents.Dr.Shi is an associate editor for IEEE Transactions on Signal Processing.He was awarded Golden Medal at the 46th International Exhibition of Inventions of Geneva in 2018,and also was the recipient of the First Prize of Science and Technology Award from China Institute of Communications in 2017,the National Excellent Doctoral Dissertation Nomination Award in 2013,the Shanghai Excellent Doctorial Dissertation Award in 2012, and the Best Paper Award from the IEEE PIMRC’09 conference.|Jingran Lin received his B.S. degree in computer communication from University of Electronic Science and Technology of China(UESTC),Chengdu,China, in 2001, and his M.S. and Ph.D. degrees in signal and information processing from UESTC in 2005 and 2007,respectively. After his graduation in July 2007, he joined the School of Communication and Information Engineering,UESTC,where he is currently an associate professor.From January 2012 to January 2013, he was a visiting scholar with University of Minnesota(Twin Cities), Minneapolis,MN,USA.|His research interests include the design and analysis of efficient optimization algorithms for the signal processing problems arising from modern communication systems.

在线阅读

2

PDF下载

172

摘要：

DOA估计是无人机应用中的一个重要任务。然而，城市复杂的电磁传播环境会降低传统基于模型驱动的DOA估计方法的性能。为了减轻环境影响，提出一种基于深度学习的DOA估计方法。具体而言，该方法采用复值卷积神经网络，以便更好的表征无人机无线信号的复包络特性。本文使用复数域概率来构建复值卷积神经网络映射函数，并进一步分析了影响复值卷积神经网络收敛性的因素。数值仿真的结果表明，针对DOA估计任务，复值卷积神经网络在收敛速度、准确性和鲁棒性等方面都优于传统的实值卷积神经网络。

Abstract:

Abstract—Direction-of-arrival (DOA) estimation is an important task in many unmanned aerial vehicle (UAV) applications.However, the complicated electromagnetic wave propagation in urban environments substantially deteriorates the performance of many conventional model-driven DOA estimation approaches.To alleviate this, a deep learning based DOA estimation approach is proposed in this paper.Specifically, a complex-valued convolutional neural network (CCNN) is designed to fit the electromagnetic UAV signal with complex envelope better.In the CCNN design,we construct some mapping functions using quantum probabilities, and further analyze some factors which may impact the convergence of complex-valued neural networks.Numerical simulations show that the proposed CCNN converges faster than the real convolutional neural network, and the DOA estimation result is more accurate and robust.

 复卷积神经网络设计及其在城市环境下无人机方向估计中的应用

DOA估计是无人机应用中的一个重要任务。然而，城市复杂的电磁传播环境会降低传统基于模型驱动的DOA估计方法的性能。为了减轻环境影响，提出一种基于深度学习的DOA估计方法。具体而言，该方法采用复值卷积神经网络，以便更好的表征无人机无线信号的复包络特性。本文使用复数域概率来构建复值卷积神经网络映射函数，并进一步分析了影响复值卷积神经网络收敛性的因素。数值仿真的结果表明，针对DOA估计任务，复值卷积神经网络在收敛速度、准确性和鲁棒性等方面都优于传统的实值卷积神经网络。

Key words: direction-of-arrival (DOA) estimation, complex-valued convolutional neural network (CCNN), unmanned aerial vehicle(UAV)

No Suggested Reading articles found!