An auto-configurable, and, adaptable, metric-driven cluster-head organisation for hybrid multi-hop routing
Nikos Migas, William Johnston Buchanan · 2006
Multi-hop ad-hoc routing is a challenging issue, because of the dynamic network topology, and limited capabilities of resource-constrained mobile devices. This paper proposes a metric-driven clustering organisation of participating nodes which provides an effective way to reduce network overhead, in contrast to traditional flooding methods. This novel method creates a network backbone for data routing and location management, which is composed of the fittest nodes. The key metrics are: mobility, buffering capabilities, throughput, network state, utilisation, and battery reserves. The paper presents the overall agent-based model, and describes the adaptability of the cluster-head metrics, and the automated reconfigurability of an ad-hoc network