GELS: Generalized Energy-Based Leader Selection in Bio-Inspired Leader‐Follower AUV Network

Abstract

Abstract An autonomous underwater vehicle (AUV) is a self-propelled, untethered underwater vehicle with minimal or no human supervision. However, the capability of a single AUV to cover large areas underwater is very limited. In this regard, a swarm of AUVs has gained significant attention in the recent years for underwater applications such as ocean exploration, mine-sweeping, surveillance, rescue missions, seabed mapping, environmental monitoring, etc. Notably, the swarm-based network can be further classified into two types, based on the architecture: cooperative multi-AUV network and leader‐follower multi-AUV network. Of these, the leader‐follower multi-AUV network with a Fixed Leader is preferable due to its simple architecture and less operational complexity. However, the inherent energy limitations of the leader in the leader‐follower model become the bottleneck for an efficient mechanism. Hence, there is a need to design a better mechanism for a multi-AUV network. In this paper, a Generalized Energy-based Leader Selection (GELS) algorithm is proposed for a multi-AUV network to increase the AUV network's transmission range and duration. Significantly, GELS can be superimposed and integrated with the existing algorithms. This proposed work integrates GELS with two bio-inspired mechanisms: Ant Colony Optimization and Particle Swarm Optimization. The dynamic leader selection using GELS results more in the travel distance and duration than the Fixed Leader network. This is because, in GELS, based on residual energy, the followers could also be allowed to lead the network; whereas, in fixed leader the followers cannot lead the network.