CADD: Collaborative adaptive duty cycle and data reduction for congestion control in LoRaWAN

The widespread adoption of Internet of Things (IoT) technologies in domains such as smart agriculture, environmental monitoring, and industrial automation has significantly increased the number of connected devices and the volume of data generated. LoRaWAN, a protocol originally designed for low-bandwidth and sporadic transmissions, faces challenges in supporting these evolving requirements. In particular, the limited channel capacity and uncoordinated ALOHA-based medium access protocol render LoRaWAN increasingly susceptible to congestion, packet collisions, and energy inefficiencies in dense deployments. This paper presents CADD, a novel channel congestion control mechanism for LoRaWAN networks that addresses end-device and gateway-level constraints while incorporating data priority. The proposed approach combines (i) a dynamic data reduction technique at the end devices to eliminate redundant transmissions and (ii) an adaptive duty cycle regulation mechanism that adjusts transmission opportunities based on channel congestion observed at the gateway. This joint strategy enables efficient use of the communication channel while ensuring the reliable delivery of critical data. The proposed method is evaluated using realistic datasets in large-scale network scenarios. Experimental results demonstrate that CADDeffectively reduces transmission frequency and energy consumption, mitigates channel congestion, and improves overall network performance and scalability in high-density IoT deployments. In scenarios with up to 500 devices, CADDachieves up to a 4× higher PDR and reduces energy consumption by approximately 54% compared to a state-of-the-art congestion control scheme and a static duty cycle baseline.