• DocumentCode
    20617
  • Title

    Understanding channel selection dynamics in dense Wi-Fi networks

  • Author

    Baid, Akash ; Raychaudhuri, Dipankar

  • Volume
    53
  • Issue
    1
  • fYear
    2015
  • fDate
    Jan-15
  • Firstpage
    110
  • Lastpage
    117
  • Abstract
    This paper aims to explain and analyze a growing problem in dense-urban wireless networks, that of co-existence between low-cost residential access points (APs) and actively-managed service provider APs in overlapping spatial, frequency, and time domains. Through detailed simulations and testbed experiments, the impact of increasing density of highly-adaptive service provider APs on the performance of typical residential APs is measured in terms of their respective channel assignment schemes. Simulation results with dense deployment of up to 500 APs/sq.km. show the benefits of centralized channel assignments, even in the presence of independent APs. In addition, it is shown that for a fixed AP density, an increase in the percentage of AP under the centralized scheme results in an increase in the throughput of surrounding independent APs. The broader implications of the simulation findings are discussed in order to develop a better macro-level understanding of dense Wi-Fi networks.
  • Keywords
    channel allocation; frequency-domain analysis; time-domain analysis; wireless LAN; actively-managed service provider; centralized channel assignment scheme; channel selection dynamics; dense Wi-Fi networks; dense-urban wireless networks; fixed AP density; frequency domains; highly-adaptive service provider density; independent APs; low-cost residential access points; overlapping spatial domains; time domains; Approximation methods; Channel allocation; Computational modeling; Dense estimation; IEEE 802.11 Standards; Mathematical model; Throughput; Wireless networks;
  • fLanguage
    English
  • Journal_Title
    Communications Magazine, IEEE
  • Publisher
    ieee
  • ISSN
    0163-6804
  • Type

    jour

  • DOI
    10.1109/MCOM.2015.7010523
  • Filename
    7010523