Online reconfiguration of channel assignment in Multi-Channel Multi-Radio wireless mesh networks

  • Authors:
  • A. Antony Franklin;Athula Balachandran;C. Siva Ram Murthy

  • Affiliations:
  • Cognitive Radio Technology Research Team, Electronics and Telecommunications Research Institute (ETRI), Daejeon 305-700, South Korea;School of Computer Science, Carnegie Mellon University, Pittsburgh, PA 15213, USA;Department of Computer Science and Engineering, Indian Institute of Technology Madras, Chennai 600036, India

  • Venue:
  • Computer Communications
  • Year:
  • 2012

Quantified Score

Hi-index 0.24

Visualization

Abstract

Efficient utilization of Multi Channel-Multi Radio (MC-MR) wireless mesh networks (WMNs) can be achieved only by intelligent channel assignment (CA) and Link Scheduling (LS). Due to the dynamic nature of traffic demand in WMNs, the CA has to be reconfigured whenever traffic demand changes, in order to achieve maximum throughput in the network. The reconfiguration of CA requires channel switching at radios which leads to disruption of ongoing traffic in the network. So, we have to consider this traffic disruption overhead while reconfiguring the network for traffic adaptation. The existing CA algorithms for MC-MR WMNs in the literature do not consider the reconfiguration overhead caused by the channel switching. In this paper, we propose a novel reconfiguration model that considers both network throughput and reconfiguration overhead to quantitatively evaluate a reconfiguration algorithm. Based on the reconfiguration model, we formulate the problem of reconfiguration of CA as a Mixed Integer Linear Program (MILP). We also propose an on-line heuristic algorithm for CA called Demand based State Aware channel Reconfiguration Algorithm (DeSARA) that finds the CA for the current traffic demand by considering the existing CA of the network to minimize the reconfiguration overhead. Through extensive simulations, we show the importance of considering the overhead in reconfiguration of CA, by comparing the performance of DeSARA with a static CA and a fully dynamic CA that does not consider the reconfiguration overhead. We also study the performance of the proposed algorithm with real network traces collected in a campus network to show its practicality.