• Login
    View Item 
    •   SMARTech Home
    • Georgia Tech Theses and Dissertations
    • Georgia Tech Theses and Dissertations
    • View Item
    •   SMARTech Home
    • Georgia Tech Theses and Dissertations
    • Georgia Tech Theses and Dissertations
    • View Item
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    Deep Learning based Intelligent Wireless Communication System

    Thumbnail
    View/Open
    YE-DISSERTATION-2020.pdf (10.76Mb)
    Date
    2020-12-10
    Author
    Ye, Hao
    Metadata
    Show full item record
    Abstract
    The objective of the thesis is to exploit the expression power of machine learning to develop, analyze, and test novel physical and medium access control (MAC) layer designs for intelligent wireless communication systems. In the physical layer, a deep learning approach is developed for joint channel estimation and symbol detection in an orthogonal frequency-division multiplexing (OFDM) system. The deep learning based approach estimates channel state information (CSI) implicitly and recovers the transmitted symbols directly. In addition, the deep learning based approach can address channel distortion in a data-driven manner. Besides exploiting deep learning to enhance the conventional communication blocks, deep neural networks (DNNs) can also be used to build a novel end-to-end wireless communication system, in which DNNs are employed to perform several key functions, including encoding, decoding, modulation, and demodulation. We develop a conditional generative adversarial net (GAN) to represent channel effects and to bridge the transmitter DNN and the receiver DNN so that the gradients for training the transmitter DNN can be obtained. Furthermore, instead of using pilot information to implicitly or explicitly estimate the unknown channel parameters in end-to-end systems, the transmitter DNN can learn to encode the input data in a way that is robust to various channel conditions without any pilots. In the MAC layer, a deep reinforcement learning based distributed resource allocation framework for vehicular communication networks is developed. An autonomous “agent”, a vehicle-to-vehicle (V2V) link or a vehicle, makes its decisions to find the optimal sub-band and power level for transmission without requiring or having to wait for global information. In addition, we also investigate the deep learning enabled over-the-air computation for efficient information sharing among the edge devices, where both the pre-processing and post-processing functions of the over-the-air computation are represented by DNNs.
    URI
    http://hdl.handle.net/1853/64179
    Collections
    • Georgia Tech Theses and Dissertations [23877]
    • School of Electrical and Computer Engineering Theses and Dissertations [3381]

    Browse

    All of SMARTechCommunities & CollectionsDatesAuthorsTitlesSubjectsTypesThis CollectionDatesAuthorsTitlesSubjectsTypes

    My SMARTech

    Login

    Statistics

    View Usage StatisticsView Google Analytics Statistics
    facebook instagram twitter youtube
    • My Account
    • Contact us
    • Directory
    • Campus Map
    • Support/Give
    • Library Accessibility
      • About SMARTech
      • SMARTech Terms of Use
    Georgia Tech Library266 4th Street NW, Atlanta, GA 30332
    404.894.4500
    • Emergency Information
    • Legal and Privacy Information
    • Human Trafficking Notice
    • Accessibility
    • Accountability
    • Accreditation
    • Employment
    © 2020 Georgia Institute of Technology