Centrality Aware AODV Routing for FANETs

Abstract

End-to-end unmanned aerial vehicle networks suffer from overloading of central nodes due to shortest-path routing strategies. This study presents the CA-AODV approach, which integrates distributed eigenvector centrality computation into the AODV routing protocol. The proposed protocol employs a two-phase eigenvector computation mechanism, RREQ (Route Request) buffering for best-path selection, and a centrality-based link cost function to divert traffic away from central nodes. Simulation results conducted with OMNeT++ on a 50-node UAV network demonstrate that CA-AODV achieves up to %3.7 improvement in packet delivery ratio, up to %25.3 reduction in end-to-end delay, and %3.5 increase in sum rate compared to standard AODV under high traffic loads.