Coverage and connectivity aware node deployment using improved NSGA-II and MOSSA approach for UWSNs

Abstract Underwater wireless sensor networks (UWSNs) are extensively utilized for monitoring of aquatic environments, oceanographic data gathering, target tracking, and disaster early warning, but their effectiveness depends on maintaining adequate sensing coverage and end-to-end connectivity in the presence of node displacement caused by ocean currents, mobility, natural disasters, and environmental interference. Under such conditions, deployed nodes drift from their intended positions and become trapped in regions where they can neither cover the area of interest nor reach the surface infrastructure, producing both interior coverage holes inside the monitored volume and boundary holes at its periphery. Most existing schemes consider only interior coverage holes and do not address boundary holes that arise after natural disasters or environmental disturbance, and most are single-objective formulations that do not jointly account for coverage, connectivity, and energy. To mitigate these gaps, a unified two-stage scheme is proposed. In the first stage, a hole-detection method based on layered delaunay triangulation (DT) and sensing-circle intersection points is used to enumerate both interior and boundary holes within the monitored volume. In the second stage, additional autonomous-underwater-vehicle (AUV) based mobile nodes are deployed at optimal positions using two complementary multi-objective metaheuristics: an improved Non-dominated Sorting Genetic Algorithm-II (NSGA-II) with multi-point crossover and a probability-based adaptive mutation operator, and a Multi-Objective Salp Swarm Algorithm (MOSSA) augmented with non-dominated sorting. A novel multi-objective fitness function is optimized that simultaneously maximises coverage quality, connectivity, and network lifetime, evaluated under a realistic underwater acoustic channel model and a Meandering Current Mobility model. The proposed scheme additionally exploits the geometric structure of overlapping coverage holes to deploy a single mobile node to repair multiple overlapping holes whenever feasible, reducing the number of additional AUVs required. The two algorithms are compared against published baselines SHORT (spherical hole-repair technique), DABVF (distributed node deployment algorithm based on virtual forces), and improved NSGA-II. MOSSA achieves a higher coverage quality, reduces average per-node energy, and extends network lifetime compared to existing schemes. Averaged over 30 independent runs, MOSSA improves coverage quality by 2.4–19.3%, connectivity by 4.5–23.8%, convergence fitness by 14.0–46.7%, and energy efficiency by 23.5–43.2% compared with improved NSGA-II, DABVF and SHORT, where each range covers the three baselines from the strongest to the weakest. In the healing scenario, the additional AUVs increase coverage from 78 to 100% by repairing both interior and boundary holes. All pairwise differences are statistically significant at p < 0.05 using the Wilcoxon signed-rank test with Holm correction.