Cross-layer Framework for Energy Harvesting-LPWAN Resource Management based on Fuzzy Cognitive Maps and Adaptive Glowworm Swarm Optimization for Smart Forest

Modern forestry research and management increasingly rely on precise environmental data. Presently, Low Power Wide Area Networks (LPWANs) offer potential advantages for such field monitoring tasks. However, their applicability requires enhancements in aspects such as power consumption, transmission range, data rate, and consistent quality of service. This paper introduces a novel control model emphasizing cross-layer collaboration, aiming to bolster the efficiency and reliability of Energy Harvesting (EH) LPWANs within the context of intelligent forest management. By employing the influential factors of EH-LPWAN as conceptual nodes, an innovative fuzzy cognitive map (FCM) can be designed. The interrelations among these concepts become instrumental in developing the cross-layer optimization model, addressing various objectives and tackling overlapping constraints. To further refine the model's efficacy, an adaptive glowworm swarm optimization (AGSO) driven dynamic FCM method is presented to ascertain the conceptual weights while facilitating real-time updates. Preliminary results manifest a noteworthy enhancement in communication range by 40.2%, a betterment in packet delivery accuracy by 19%, and an extension in the LoRaWAN's projected lifespan by 33.8% during scenarios with diminished EH rates. It's evident that the energy self-sustainability of EH nodes coupled with the data handling capacity of the entire network fully aligns with the stringent real-time and consistency criteria mandated for meticulous forest observation.