Wildfire monitoring systems deployed in remote, suburban, and forested areas must operate under stringent constraints such as limited energy resources, device heterogeneity, and unreliable infrastructure. Traditional minimum spanning tree algorithms fall short in supporting such complex, distributed applications. In this work, we present a novel framework for optimizing peer-to-peer Internet of Things (IoT) networks tailored to environmental monitoring scenarios, such as real-time wildfire detection.Our approach generalizes classical connectivity algorithms by categorizing devices into three roles: weak devices requiring connection, mandatory devices with limited capacity, and discretionary devices that enhance performance. The proposed objective function jointly minimizes connection costs, capacity violations, and network fragmentation, enabling adaptive trade-offs that improve feasibility and performance in dynamic environments.To demonstrate practical viability, we implement our model using WebRTC technology, enabling direct peer-to-peer communication among IoT nodes. This architecture supports low-latency, decentralized data exchange without reliance on centralized cloud infrastructure, crucial for time-sensitive wildfire alerts. The system is integrated with the Stack4Things (S4T) framework, showcasing energy-efficient, scalable deployment of distributed applications in complex IoT environments.
An Adaptive and Resilience-Preserving Peer-to-Peer IoT Network for Wildfire Monitoring
Garofalo, Marco;Tricomi, Giuseppe
;De Meo, Pasquale;Merlino, Giovanni;Longo, Francesco;Puliafito, Antonio
2025-01-01
Abstract
Wildfire monitoring systems deployed in remote, suburban, and forested areas must operate under stringent constraints such as limited energy resources, device heterogeneity, and unreliable infrastructure. Traditional minimum spanning tree algorithms fall short in supporting such complex, distributed applications. In this work, we present a novel framework for optimizing peer-to-peer Internet of Things (IoT) networks tailored to environmental monitoring scenarios, such as real-time wildfire detection.Our approach generalizes classical connectivity algorithms by categorizing devices into three roles: weak devices requiring connection, mandatory devices with limited capacity, and discretionary devices that enhance performance. The proposed objective function jointly minimizes connection costs, capacity violations, and network fragmentation, enabling adaptive trade-offs that improve feasibility and performance in dynamic environments.To demonstrate practical viability, we implement our model using WebRTC technology, enabling direct peer-to-peer communication among IoT nodes. This architecture supports low-latency, decentralized data exchange without reliance on centralized cloud infrastructure, crucial for time-sensitive wildfire alerts. The system is integrated with the Stack4Things (S4T) framework, showcasing energy-efficient, scalable deployment of distributed applications in complex IoT environments.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


