Controlling Adversarial Pheromone-Based Infections via Quarantine Strategies in Foraging Robot Swarms

Virtual pheromone trails that facilitate efficient and adaptable coordination among foraging robot swarms are vulnerable to threats that exploit stigmergic communication. This study investigates the impact of a fake resource attack on the performance of a pheromone-based foraging algorithm and demonstrates the effectiveness of a “quarantine strategy” in mitigating the attack. The study simulates the fake resources and examines the swarm's behavior as robots are attracted to these fake resource locations. To prevent access to fake resources, circular quarantine regions are deployed, and a distance-based merging algorithm is implemented to reduce storage require-ments. The experiments are conducted with varying numbers of fake resources and simulation time using the ARGoS simulation environment. The study illustrates the impact of pheromone trail exploitation in pheromone-based foraging algorithms and an effective defense mechanism for such scenarios. The results show a significant decrease in the foraging algorithm's performance with an increase in the number of fake resources. Additionally, the study demonstrates the effectiveness of the quarantine strategies in reducing the collection of fake resources. Overall, this research highlights the fragility of pheromone-based foraging algorithms and provides a defense mechanism to protect against attacks exploiting these systems. The study's findings can inform the development of more robust and efficient foraging algorithms that are resilient to attacks for swarm robotics systems in real-world scenarios.