Low Power Wide Area Network (LPWAN) technologies are ideal for collecting data in remote locations with challenging propagation environments. LoRaWAN, which uses the LoRa radio modulation, is considered the leading LPWAN technology. However, LoRa’s communication range often decreases in real-world deployments due to attenuation and fading effects, resulting in packet losses and errors. In rural areas, for instance, the performance is highly dependent on the terrain topology, where a mountain can create a shadow zone. Although adding gateways can improve coverage, it is not always feasible due to installation costs or location constraints.

Many works address blind spots with multi-hop solutions that introduce high traffic overhead and result in high energy consumption. A more lightweight alternative is for disconnected nodes (DNs) to discover neighbouring nodes (Relay Nodes, RNs) that can reach a gateway (GW). However, RNs must periodically wake up to relay packets from DNs to the GW, increasing their energy consumption. In remote deployments, frequent maintenance is impractical. Thus, an energy-efficient relaying mechanism design is essential to reduce energy consumption and minimize battery replacements. To this end, we propose SFrelay, a novel energy-efficient relaying mechanism to mitigate blind spots in LoRa-based networks. SFrelay maintains peer-to-peer synchronization between DNs and RNs, offering sufficient communication opportunities to ensure a good packet delivery ratio. Through an optimal balance between synchronization and communication parameters, SFrelay outperforms existing solutions in terms of energy consumption, mainly in scenarios with long transmission periods. The results have been validated through analytical modelling, simulation, and testbed experiments.