Hsieh, Ping-Chun (2018-08). Network Algorithms for Control and Communication for IoT Applications. Doctoral Dissertation. Thesis uri icon

abstract

  • Internet of Things (IoT) technology is accelerating the integration of wireless communication networks and physical systems, such as the emerging networked transportation systems and industrial automation applications. By leveraging IoT, these physical systems are envisioned to perform critical tasks more reliably and efficiently based on the real-time control information provided by various types of sensors. To achieve this vision, there are many fundamental challenges to be tackled in both theory and implementation for control and communication of these physical systems. Specifically, we study the design and implementation of network algorithms for the following IoT applications: _ We develop scheduling schemes for networked transportation systems. Different from the conventional scheduling problem in computer networks, we consider practical constraints of the physical systems, such as switch-over delay, estimation errors, finite buffer sizes, and partially-connected systems and propose a throughput-optimal scheduling policy. _ We develop wireless network algorithms for collecting and disseminating critical control information. We design distributed algorithms for real-time wireless ad hoc networks. Moreover, we design scheduling algorithms for optimizing Quality of Experience for video delivery applications by applying diffusion approximation. _ We develop a low-latency wireless testbed for prototyping real-time wireless scheduling policies as well as the proposed network algorithms.
  • Internet of Things (IoT) technology is accelerating the integration of wireless communication
    networks and physical systems, such as the emerging networked transportation
    systems and industrial automation applications. By leveraging IoT, these
    physical systems are envisioned to perform critical tasks more reliably and efficiently
    based on the real-time control information provided by various types of sensors. To
    achieve this vision, there are many fundamental challenges to be tackled in both theory
    and implementation for control and communication of these physical systems.
    Specifically, we study the design and implementation of network algorithms for the
    following IoT applications:
    _ We develop scheduling schemes for networked transportation systems. Different
    from the conventional scheduling problem in computer networks, we consider
    practical constraints of the physical systems, such as switch-over delay, estimation
    errors, finite buffer sizes, and partially-connected systems and propose a
    throughput-optimal scheduling policy.
    _ We develop wireless network algorithms for collecting and disseminating critical
    control information. We design distributed algorithms for real-time wireless
    ad hoc networks. Moreover, we design scheduling algorithms for optimizing
    Quality of Experience for video delivery applications by applying diffusion approximation.
    _ We develop a low-latency wireless testbed for prototyping real-time wireless
    scheduling policies as well as the proposed network algorithms.

publication date

  • August 2018