Enabling LPWAN Co-existence via PHY-layer Packet Filtering in Dense Deployments

Abstract

LoRa has seen widespread adoption for long-range Internet of Things (IoT) applications, particularly in dense urban sensing deployments. While LoRaWAN supports multiple channels to improve scalability, practical performance is constrained by hardware limitations of commercial gateways, which can process only a limited number of concurrent packets per channel due to a fixed number of demodulator/decoder chains. In overlapping networks, this leads to severe contention for decoder resources and significant throughput degradation, exacerbated by LoRa’s long range and lack of early packet discrimination. We present LoRaPHYID, a PHY-layer filtering framework that enables device–gateway association prior to decoding. By leveraging the LoRa PHY structure, LoRaPHYID allows gateways to selectively process only intended packets and discard others early, effectively isolating coexisting networks without requiring additional spectrum. We implement LoRaPHYID on SDR-based gateways and evaluate it using a real-world testbed with commercial LoRa devices. Results show up to 2× improvement in per-gateway throughput under dense, fully overlapping traffic, demonstrating the effectiveness of PHY-layer isolation in mitigating decoder contention.