Abstract

To mimic the operations in fixed infrastructures and to solve the routing scalability problem in large ‘mobile ad hoc networks’, forming clusters of nodes has been proven to be a promising approach. However, when existing weighted clustering algorithms calculate each node’s weight, they either consider only one metric or rely on some metrics collected from extra devices. This often leads to a higher rate of re-clustering. This article presents a robust weighted clustering algorithm called FMAC (fuzzy controlled multi-hop adaptive clustering), to form and maintain more stable clusters. In FMAC, the clusterheads are elected based on energy of nodes, rate of energy depletion, relative mobility with its neighbours, radio-range and cardinality of the set of neighbours. Two fuzzy controllers are embedded in each node for selecting clusterheads and monitoring join of nodes to clusters. Simulation results show that FMAC prolongs lifetime of ad hoc networks and has a lower clusterhead change rate and re-affiliation rate than other existing algorithms.

Key words: Adaptive, ad hoc networks, clustering, fuzzy controller, protocol.