Realistic Traffic Modeling and Performance Evaluation of a Blockchain-Enabled LoRaWAN

In the Internet of Things (IoT) scenario, two of the most important innovations in recent years are Edge Computing and Low Power technologies, like LoRaWAN. Edge Computing facilitates computations to be executed in proximity to users, delivering advantages such as reduced response times, optimized bandwidth usage, and enhanced scalability for IoT applications. On the other hand, LoRaWAN stands out as an IoT communication technology engineered for secure, low data-rate transmissions with high energy efficiency. However, implementing the edge computing paradigm into LoRaWAN presents challenges due to its centralized cloud-based architecture, which conflicts with the principles of edge computing. In previous years, proposals have emerged to decentralize LoRaWAN and integrate it at the edge level. However, these proposals often result in the creation of new protocols that diverge from the standard LoRaWAN framework and are not backward compatible. Furthermore, existing proposals are typically tested on arbitrary testbeds and traffic conditions, failing to account for the diverse applications and traffic characteristics inherent in real-world scenarios. In this study, we enhance and refine DeLoRaN, our system architecture facilitating the decentralized operation of LoRaWannetworks at the edge layer, where network control operations are executed at the gateway level. Additionally, we develop and present a data-driven traffic model for LoRaWAN that reflects real-world scenarios, derived from extensive capturing of LoRaWantraffic over several months. This model leverages packets from devices across multiple networks and serves diverse applications not under our direct control. Using this model, we validate the performance of DeLoRaN by simulating device traffic generation based on our data-driven realistic model.