Swarm intelligence for hole detection and healing in wireless sensor networks

The increasing demand for wireless sensor networks to monitor specific regions has prompted extensive research on sustaining coverage over time. The main threat to this goal arises from coverage holes caused by random node deployment or failures. This study proposes a swarm intelligence-based algorithm to detect and heal coverage holes. The swarm of agents relies on local and relative information, activating in response to detected holes and navigating a potential field toward the closest hole. The agents quantize their perceptions to disperse efficiently, approaching holes from different directions to accelerate healing. Based on geometric criteria, the swarm deploys at locally optimal positions along hole borders while preventing redundant deployments. Agents deployment update the potential field, guiding the rest of the swarm toward unhealed areas and ensuring dynamic detection and tracking of new holes, even near the region frontier. Experimental studies demonstrate superior coverage restoration compared to state-of-the-art solutions, showing good scalability and flexibility to different hole sizes, shapes, and multiplicity. Moreover, it exhibits high robustness to the corruption of agents’ perceptions and to their failure, while efficiently managing the battery level.