Usage
  • 13 views
  • 7 downloads

A SIMULATION-DRIVEN APPROACH TO SCHEDULING IN DUTY-CYCLED NETWORKS

  • Author / Creator
    Ho, Van H
  • Transceiver duty-cycling (DC) is a popular technique to conserve energy in a wireless sensor network (WSN). In this thesis, our overall objective is to study the performance of a DC WSN. Namely, we consider the performance of a DC WSN from the point of throughput, as well as energy consumption, under deterministic DC behavior. Our elementary notion of traffic is that of a traffic flow, routed via multiple hops from a source to a destination node. The traffic flows are, by default, assumed to be greedy, i.e., capable of making use of as much rate as they are given. We use max-min fairness as our criterion for allocating the available capacity across the multiple flows. Towards this end, we explore the following research questions. A first question is that of relating duty-cycling and achievable throughput through the construction of deterministic transmission schedules, i.e., by ensuring coordinated medium access and hence avoiding collisions and avoiding idle listening. A second question is that of throughput improvement in DC WSN by means of employing network coding (NC), again via the construction of a deterministic transmission schedule. The key component of our methodology is a modelling step whereby the time-varying topologies of DC WSNs are captured by the notion of repeating “stages”, i.e., time durations during which the topology is constant. The periodically repeating stages allow us to express optimization objectives for the throughput of the traffic flows. Whereas the optimization enables the determination of the per-flow rates, a novel simulation-based approach is introduced enabling the construction of periodic TDMA transmission schedules achieving, or at least closely approximating, those per-flow rates. The schedule construction process is based on the conjecture that the periodic behavior of a DC WSN is bound to result in a periodic steady state behavior for the network. Consequently we use simulation to extricate, and use, the periodic scheduling induced by the periodic steady state of the network. Through numerous simulation studies and in comparison with the per-flow rates produced by the optimization step, we demonstrate that the technique is accurate and applicable to arbitrary topology DC WSNs. We also demonstrate how this approach can be used in combination with NC as well. While the particular NC scheme is one of numerous possible, it is sufficient to demonstrate the technique’s advantages. The thesis concludes by outlining how the basic idea can be extended to DC WSNs with different DC periods, non-greedy traffic demands, and variations of NC schemes.

  • Subjects / Keywords
  • Graduation date
    2015-11
  • Type of Item
    Thesis
  • Degree
    Doctor of Philosophy
  • DOI
    https://doi.org/10.7939/R3N010004
  • License
    This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.
  • Language
    English
  • Institution
    University of Alberta
  • Degree level
    Doctoral
  • Department
    • Department of Computing Science
  • Supervisor / co-supervisor and their department(s)
    • Nikolaidis, Ioanis (Computing Science)
  • Examining committee members and their departments
    • Liscano, Ramiro (Univerisity of Ontario Institute of Technology)
    • Jiang, Hai (Electrical & Computer Engineering)
    • Lu, Paul (Computing Science)
    • Harms, Janelle (Computing Science)
    • Nikolaidis, Ioanis (Computing Science)