Biomimetic Coverage Optimization in WSNs via an Enhanced Spider-Wasp Optimizer with Adaptive Learning and Dual-Median-Point Guidance

Coverage optimization in wireless sensor networks (WSNs) is a non-convex deployment problem on which population-based optimizers often converge prematurely and lose diversity. This paper proposes the Enhanced Spider-Wasp Optimizer (ESWO), an extension of the Spider-Wasp Optimizer (SWO) conceptually inspired by the foraging and nesting behavior of Pompilidae wasps. ESWO integrates (i) an Adaptive Learning Strategy (ALS) that uses four types of positional gaps among individuals to sustain diversity, (ii) a dual-median-point guidance (DMPG) mechanism that steers exploitation towards the midpoint of two elite solutions, and (iii) a position-update rule with nest-overlap avoidance. Coverage is evaluated on a discretized target grid with a Euclidean binary-disk sensing model and an energy-penalized fitness. In four scenarios (25–65 nodes; 50 × 50 to 165 × 165 m2), ESWO was compared with seven algorithms, including SWO. It achieved coverage rates of 81%, 88%, 92% and 85%, the highest in every scenario (tied with MGWO at 81% in the smallest). In the 100 × 100 m2 scenario it reached 85% coverage after 523 iterations, against 601–698 for the competitors, although final convergence iteration counts were not uniformly the lowest. Over 15 runs, ESWO obtained a mean coverage of 90.2% (SD 1.2), a 95% success rate and the lowest estimated energy consumption (175.3 J).