Many optimisation methods for wireless sensor networks are tested on large scenarios where the true best solution is unknown. In contrast, this article deliberately uses a small‑to‑medium test case where all possible sensor configurations can be enumerated, so the global optimum can be computed with a brute‑force search. This optimum then serves as a ground‑truth benchmark to rigorously evaluate the proposed genetic algorithm.

The researchers show that their constraint‑aware genetic algorithm consistently reaches solutions that are very close to this theoretical optimum, but with far lower computational cost than exhaustive search. They systematically study how population size, crossover strategy, mutation probability and constraint‑handling affect convergence, demonstrating that the algorithm can deliver high‑quality configurations quickly.

In UnderSec, where optimisation may need to run on edge devices with limited computing power, this balance between solution quality and execution time is crucial. It allows sensor networks to be re‑optimised in near real time as threats, interference or environmental conditions evolve around underwater and other critical infrastructures.