Secure Split and Classification Mechanism for LoRa Networks Against Active and Reactive Interference Attacks

In the Internet of Things (IoT) domain, resource-constrained devices are highly susceptible to malicious interference due to limited duty cycles, finite energy reserves, and an inability to perform clear channel assessment. These vulnerabilities are further exacerbated in low-power wide-area networks (LPWANs), where the absence of centralized coordination renders the network an attractive target for jamming attacks. In this work, we introduce a novel split and classification mechanism designed to enhance LoRa packet delivery under both constant and reactive jamming conditions. Our approach partitions each LoRa frame into sub-packets and applies a realtime classification algorithm to identify and mitigate corrupted segments, thereby improving the effective data extraction ratio (DER). We validate the mechanism through extensive simulated scenarios across diverse network densities. Comparative analysis against the standard LoRa protocol demonstrates a significant increase in packet success rate and commensurate gain in DER, without incurring additional energy overhead.