Non-Euclidian geographic routing in wireless networks

  • Authors:
  • Niklas Carlsson;Derek L. Eager

  • Affiliations:
  • Department of Computer Science, University of Saskatchewan, Saskatoon, SK, Canada S7N 5C9;Department of Computer Science, University of Saskatchewan, Saskatoon, SK, Canada S7N 5C9

  • Venue:
  • Ad Hoc Networks
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Greedy geographic routing is attractive for large multi-hop wireless networks because of its simple and distributed operation. However, it may easily result in dead ends or hotspots when routing in a network with obstacles (regions without sufficient connectivity to forward messages). In this paper, we propose a distributed routing algorithm that combines greedy geographic routing with two non-Euclidian distance metrics, chosen so as to provide load balanced routing around obstacles and hotspots. The first metric, Local Shortest Path, is used to achieve high probability of progress, while the second metric, Weighted Distance Gain, is used to select a desirable node among those that provide progress. The proposed Load Balanced Local Shortest Path (LBLSP) routing algorithm provides loop freedom, guarantees delivery when a path exists, is able to efficiently route around obstacles, and provides good load balancing.