A comparative analysis of metaheuristic-based clustering schemes for improving the network lifetime in flying ad hoc networks

Abstract

Communication among the unmanned aerial vehicles (UAVs) in flying ad hoc network (FANET) is a vital design aspect. This is ascribed to the highly dynamic nature of the UAVs, along with the constraints in the battery resources encountered. Devising a technique that can improve the efficiency in routing along with a stable topology in FANETs is essential. In order to do this, the paper attempts to provide a comparative analysis of two different clustering methodologies for improving the lifetime of operation of FANETs. The paper implements a clustering methodology, which employs a hyper heuristic method for selecting optimal clusters and cluster heads (CHs) using glowworm swarm optimisation (GSO) and firefly algorithm (FA). Secondly, a hybrid algorithm based on particle swarm optimisation (PSO) and firefly algorithm (FA) is applied. Connectivity, distance, energy, and neighbourhood degree are the key factors considered for the optimal selection purpose. Extensive simulations were carried out over different network areas and node densities to evaluate and compare the performances of the methods. The evaluation was based on the cluster building time (CBT), energy consumption by the network, alive node analysis and overall improvement in the network lifetime. Results largely validated the better performance of the hybrid PSOFA-based clustering scheme.