Scented Node Protocol for MANET Routing

Abstract

Ant Colony Optimization (ACO) for biologically inspired networking introduces performance gains over classical routing solutions for Mobile Ad Hoc Networks (MANETs). However, the current ACO protocols involve significant amount of overhead and do not fully reflect the wireless interference effects in routing decisions. In ant routing, sources send out ant-based control packets for route discovery and path maintenance. Destinations can assist ant packets by disseminating scent messages to provide better guidance for route discovery and thus effectively reduce the protocol overhead. For that purpose, Scented Node Protocol (SNP) is introduced for interference-aware routing with novel scent diffusion and reinforcement mechanisms. The wireless link rates are measured by identifying the node pairs that are the most impacted by wireless interference, and network flows are routed to avoid severe interference effects among concurrent wireless transmissions. The throughput and overhead performance of SNP is evaluated through extensive realistic simulations for dynamic MANET environment. The resulting amount of overhead for scent and ant packets is also evaluated through the asymptotic analysis of scaling laws, as the network size grows, and through the dynamic analysis of the finite overhead constraint, by discussing the possible effects of local network coding on scent dissemination between neighbor nodes. Our results verify the throughput and overhead gains of biologically inspired SNP in wireless networks over the existing ACO and MANET routing protocols. DOI: 10.4018/978-1-61350-092-7.ch013