Wireless Networks

, Volume 17, Issue 5, pp 1287–1304

Seeker: A bandwidth-based association control framework for wireless mesh networks


DOI: 10.1007/s11276-011-0349-4

Cite this article as:
Gupta, D., Mohapatra, P. & Chuah, CN. Wireless Netw (2011) 17: 1287. doi:10.1007/s11276-011-0349-4


The rapid deployment of wireless mesh networks across universities and enterprises, and the pervasiveness of mobile devices equipped with Wi-Fi connectivity, has resulted in a scenario wherein end users have the option to choose from a multitude of access points at any given location. Moreover, with the increasing availability of rich online content, there has been a steady increase in mobile Internet traffic. Since the choice of access point that a user associates with will directly impact his performance, it is imperative that there exist an efficient association control mechanism, in order to enhance the end user’s experience. As part of this work, we propose Seeker, a novel framework for association control in wireless networks that utilizes “available bandwidth” as the design metric. The goal of Seeker is to assist the mesh network in making an intelligent decision regarding which access point a client should associate with. As part of our scheme, we implement and evaluate a passive tool to estimate available bandwidth in wireless networks. We then describe how we use this tool to implement our association control scheme, and evaluate it via extensive experiments on an outdoor testbed. Seeker takes into consideration the performance of the mesh backhaul, in addition to the client-to-AP link quality, thereby achieving significant advantages over traditional association control schemes for wireless-LANs.


Wireless mesh Available bandwidth Association control 

Copyright information

© Springer Science+Business Media, LLC 2011

Authors and Affiliations

  • Dhruv Gupta
    • 1
  • Prasant Mohapatra
    • 1
  • Chen-Nee Chuah
    • 2
  1. 1.Department of Computer ScienceUniversity of California DavisDavisUSA
  2. 2.Department of Electrical EngineeringUniversity of California DavisDavisUSA

Personalised recommendations