A novel clustering method for fault recovery and routing in mobile ad-hoc networks

K. B. Gurumoorthy, S. Allwin Devaraj, S. Gopinath, Tanweer Ali

Research output: Contribution to journalArticlepeer-review

6 Citations (Scopus)


Mobile ad-hoc network (MANET) is a group of self-organized autonomous wireless devices that serve communication in human unattended and emergency environments. The network is decentralized and uses wireless links for communication, which is vulnerable to network resource depletion rapidly. Energy and link stability are vital factors that support the prolonged operation of the network, obstructing earlier resource depletions. These depletions are overwhelmed with the help of scattered, isolated nodes; the process of augmenting them increases the control overhead. We propose a genetic algorithm-based routing (GAR) with fault route recovery (FRR) caused due to isolated nodes. In this method, clustering is used for energy balancing for retaining the live nodes' count reliably. The FRR phase prevents cluster head flooding using the local rerouting process. The former phase of GAR-FFR governs the network's energy optimization aiming at controlled energy consumption. The later part reduces routing overhead due to route failures, preventing backtracking to the cluster head. The proposed GAR-FFR is analyzed using the following metrics: throughput, packet delivery ratio, live nodes count, remaining energy, and routing overhead. The proposed GAR-FRR achieves 15.4% high throughput, 16.29% high live nodes, 8.9% high remaining energy, and 21.04% fewer control packets for different rounds, compared with the existing A-ECOPS and REAC-IN methods.

Original languageEnglish
JournalInternational Journal of Communication Systems
Issue number15
Publication statusPublished - 10-2021

All Science Journal Classification (ASJC) codes

  • Computer Networks and Communications
  • Electrical and Electronic Engineering


Dive into the research topics of 'A novel clustering method for fault recovery and routing in mobile ad-hoc networks'. Together they form a unique fingerprint.

Cite this