QoS-Aware Multichannel Resource Allocation in LoRaWAN Networks

The ALOHA uplink and best-effort service model as implemented by LoRaWAN is inherently ill-suited to realizing differentiated QoS as device populations and traffic heterogeneity increase. This paper proposes a QoS-aware resource-allocation framework based on multichannel parallelism in LoRaWAN networks. The key idea is to exploit the spreading factors' quasiorthogonality and the independence of frequency channels, which allow viewing the network as a pool of parallel transmission resources, and to dynamically distribute that pool across traffic classes. A proportional, priority-weighted policy allocates resources to classes based on their instantaneous weighted demand, and an efficiency-aware metric within each class biases nodes toward lower-airtime (faster) resources to improve utilization. The resulting design is still fully distributed, low-complexity, yet it consistently reduces latency and increases throughput for high-priority and high-reliability traffic without starving best-effort flows.