The research, authored by J. Maheswari of the Department of Electronics and Communication Engineering at PSRR College of Engineering, Sivakasi, and J. Senthil Kumar of Mepco Schlenk Engineering College, also in Sivakasi, Tamil Nadu, India, introduces a two-stage optimization framework. The first stage elects cluster heads using a Commutable Barnacle Optimization (CBO) algorithm, while the second stage performs routing using a population-based Botox Sailfish Optimization (BSO) algorithm. The approach was implemented and benchmarked in the NS-3 network simulator, where it was compared against several state-of-the-art protocols using a suite of performance indicators including network lifetime, energy consumption, packet delivery ratio, and end-to-end delay.

The authors describe their contribution as addressing the twin challenges of network performance reliability and energy efficiency, which they argue have limited the practical deployment of underwater sensing systems. In a conventional cluster-based UWSN, nodes are grouped into clusters, and one node in each cluster is designated as the cluster head (CH). Member nodes send their sensed data over short acoustic links to the cluster head, which aggregates the data and forwards it, often over longer distances or multi-hop paths, toward a surface sink or buoy. Because cluster heads handle the bulk of the transmission burden, they deplete their batteries far faster than ordinary member nodes. If the cluster head election is performed naively, for example by simply picking the node with the highest residual energy, the network can quickly become unbalanced, with certain regions exhausting their nodes and losing connectivity.

Maheswari and Senthil Kumar tackle this imbalance with the Commutable Barnacle Optimization algorithm, which draws its inspiration from the mating behavior of barnacles as originally formalized in the Barnacles Mating Optimizer literature. In this family of algorithms, candidate solutions are treated as individuals in a simulated population whose mating process mixes genetic material to produce offspring solutions. The key mechanism is that longer-range matings explore widely separated regions of the search space, while shorter-range matings exploit promising local regions through a process akin to local search. The researchers adapt this scheme to the cluster head election problem by encoding potential cluster head selections as candidate solutions and evaluating them against a fitness function that weighs residual energy, distance to the sink, node density, and the average intra-cluster communication distance. By iterating through generations of candidate solutions, the CBO algorithm converges on a set of cluster heads that distributes the energy load more evenly across the network, avoiding the energy holes that plague simpler election heuristics.

Once cluster heads are chosen, the data they aggregate still needs a reliable path to the surface sink, and it is here that the second algorithm takes over. The Botox Sailfish Optimization algorithm, a population-based metaheuristic, models the cooperative hunting behavior of sailfish, which herd schools of sardines to concentrate them for efficient capture. In the routing context, each sailfish in the population represents a candidate multi-hop path from a cluster head to the sink. The algorithm iteratively updates these candidate paths, balancing exploration of new routes against exploitation of the best routes found so far, with an energy-aware fitness function that penalizes paths passing through nodes with low residual energy or through long transmission hops, which in underwater acoustics are disproportionately expensive in terms of power. The authors report that combining the two algorithms yields a routing protocol in which cluster formation and path selection are jointly optimized rather than treated as independent problems, producing a more coherent balance between local data aggregation and long-distance forwarding.

The choice of underwater acoustics as the transmission medium is central to why this problem is so difficult. Terrestrial wireless sensor networks typically rely on radio frequency links, but radio waves attenuate rapidly in conductive seawater, rendering them impractical beyond short ranges. Optical links offer enormous bandwidth but suffer severe scattering, particularly in turbid coastal waters. Acoustic communication, by contrast, can propagate over kilometers, which makes it the default choice for UWSNs, but it comes with a laundry list of pathologies: propagation speeds of roughly 1,500 meters per second introduce long and variable delays, available bandwidth is typically limited to a few kilohertz to tens of kilohertz, the underwater channel exhibits strong frequency-dependent attenuation, and multipath reflections from the surface and seafloor cause fading and intersymbol interference. Each transmission at higher power or over longer distance consumes substantially more battery, so the design of the routing protocol has a direct and dramatic impact on how long the network as a whole remains operational.

Previous work in the field has explored a variety of approaches to mitigate this energy drain. Chain-based protocols organize nodes into linear chains to reduce transmission distances. Firefly-inspired protocols use dynamic mating optimization to select relay nodes. Chimp optimization, hunger games search, and hybrid cat cheetah optimization have all been applied to node clustering and multi-hop routing in recent years. Swarm intelligence approaches based on sailfish optimization have also been used for controller selection in software-defined wireless sensor networks. Maheswari and Senthil Kumar position their work within this lineage but argue that the combination of Commutable Barnacle Optimization for cluster head election and Botox Sailfish Optimization for routing offers a novel pairing that better balances exploration and exploitation at both stages of the protocol simultaneously, rather than optimizing one stage and leaving the other to a heuristic.

The evaluation in NS-3, a discrete-event network simulator widely used in academic and industrial research for protocol validation, subjects the proposed protocol to a battery of tests across varying network sizes and traffic conditions. The authors compare their approach against several existing state-of-the-art protocols and report that their combined CBO and BSO framework demonstrates superior performance across the set of inspected indicators. The paper’s graphical abstract and accompanying figures illustrate the network topology, the algorithmic flow, and the comparative results across network lifetime, residual energy distribution, packet delivery ratio, and delay. While the specific numerical improvements are detailed in the full paper behind the journal’s subscription, the authors characterize their protocol as validating the efficiency of both the CBO and BSO algorithms across the tested scenarios.

The implications of this work extend well beyond academic curiosity. The Internet of Underwater Things, a growing research area that envisions dense networks of smart underwater sensors connected to cloud infrastructure, depends on solving precisely the kind of energy management problem that this paper addresses. Applications include monitoring of oil and gas pipelines, early detection of tsunamis and undersea earthquakes, tracking of marine pollutants, precision aquaculture, military surveillance, and climate science, all of which require sensor nodes to remain operational for months or years without the possibility of battery replacement. Every percentage point of improvement in energy efficiency translates into longer mission durations and lower deployment costs, since fewer replacement nodes or recovery trips are needed. Furthermore, the paper situates underwater sensing within the broader context of cyber-physical systems and the Internet of Things, arguing that reliable underwater links are a prerequisite for extending these paradigms into the ocean.

The study also reflects a broader trend in networking research toward metaheuristic and bio-inspired algorithms as practical tools for handling the combinatorial complexity of modern protocol design. Problems such as optimal cluster head selection and optimal multi-hop routing are computationally hard; exhaustive search over all possible clusterings and routings is intractable for networks of realistic size. Metaheuristics sidestep this by navigating the search space intelligently, accepting good-enough solutions that can be computed within the time and resource budgets available to network nodes. The Barnacles Mating Optimizer, from which the CBO algorithm derives, was originally proposed in 2020 in Engineering Applications of Artificial Intelligence as a general-purpose algorithm for engineering optimization problems, and has since found applications across many domains. Similarly, sailfish-inspired optimizers have shown promise in wireless network resource allocation. Maheswari and Senthil Kumar’s contribution lies in adapting these tools to the specific constraints and cost functions of the underwater acoustic environment, rather than porting them wholesale from terrestrial contexts.

The authors declare no competing financial interests and note that the research received no external funding. Data will be made available upon reasonable request, and the corresponding author is J. Maheswari at PSRR College of Engineering. The article was received on 3 January 2025, accepted on 1 April 2026, and published on 8 May 2026 as part of Volume 17 of the journal, on pages 1027 through 1038. As the field moves toward practical large-scale deployments of underwater sensor networks, protocols of this kind, which treat energy as the scarcest and most precious resource and apply principled optimization at every layer of the communication stack, will likely play a central role in turning the promise of a fully networked ocean into operational reality.