Route Cache Based Load Balancing Scheme for Mobile Ad-Hoc Networks

  • Authors:
  • Young-Duk Kim;Jin-Wook Kim;Won-Seok Kang;Dong-Ha Lee

  • Affiliations:
  • Dept. of Advanced Industrial Science and Technology, Daegu Gyeongbuk Institute of Science and Technology, Daegu, Korea 700-742;Dept. of Advanced Industrial Science and Technology, Daegu Gyeongbuk Institute of Science and Technology, Daegu, Korea 700-742;Dept. of Advanced Industrial Science and Technology, Daegu Gyeongbuk Institute of Science and Technology, Daegu, Korea 700-742;Dept. of Advanced Industrial Science and Technology, Daegu Gyeongbuk Institute of Science and Technology, Daegu, Korea 700-742

  • Venue:
  • NBiS '08 Proceedings of the 2nd international conference on Network-Based Information Systems
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

Recently, many on demand routing protocols are suggested to support mobile ad hoc network (MANET) which is self-organized network. The most represented protocols are on-demand routing schemes such as DSR and AODV. These protocols set up routing paths by flooding Route Request (RREQ) packets during route discovery procedure and receive Route Reply (RREP) packets from destination nodes. In order to reduce route discovery latency, every node may use the previous routing information in its route cache and reply RREP instead of the destination. Although the route cache mechanism is simple and efficient approach to enhance the performance, it may result in unnecessary latency when it uses incorrect cache information. In addition, a certain node located around source can be easily congested because it replies every RREQ with RREP and all data flows are concentrated on this node. In this paper, we suggest a dynamic cache monitoring scheme and suppress RREP on heavily congested node to achieve load balancing. To do this, we have defined appropriate queue threshold values and parameters for congestion resolution. We also propose a solution for RREP storm problem which is another side effect of the route cache. Finally throughout the simulation, we have illustrated that proposed scheme shows better performance than the standard algorithm in heavily congested environments.