Obtaining Security Characteristics through Simultaneous Reception of LoRaWAN Packets in Redundant Gateways

The Long-Range Wide Area Network (LoRaWAN), which combines energy efficiency with long-range communication, is one of the most significant IoT communication technologies. In a variety of industries today, ensuring the security and safety of network traffic is a crucial concern, and security is particularly crucial as digitalization accelerates in these domains. Standard encryption methods are commonly employed to safeguard IoT network information; however, the limited resources of most IoT devices restrict the use of high-level encryption techniques. Consequently, IoT networks require practical solutions for securing their data. In this study, we develop and implement an approach to enhance LoRaWAN wireless network security using redundant data from multiple gateways. Our approach is based on the behavioral security method where we utilize a deep neural network (DNN) model to create a classifier that can differentiate between packets from the attacker and the target node using the metadata collected from different gateways. Our results show that the model can accurately differentiate between data originating from the target node and data from the attacker with a 100% success rate, as long as the attacker and the targeted node are not co-located.