Abstract
<title>Abstract</title> <p>Conventional LoRaWAN technology relies primarily on static routing based on received signal strength (RSS), which sometimes leads to severe congestion at nearby gateways and accelerates power consumption in areas with heavy traffic. This paper proposes a model for improving routing in LoRaWAN networks based on Software-Defined Networking (SDN) technology. The proposed model introduces a dynamic cost function that simultaneously analyzes the remaining battery power in end devices, geographical distance, and gateway queue length, thereby ensuring the selection of the most suitable gateway. Simulations were conducted using NS-3 with 100 end devices and 4 gateways. The results of these simulations showed that the SDN controller successfully balanced the network load, recording 19, 24, 27, 30 devices across the gateways, compared to conventional congestion models that are concentrated at a single point. Among the results, we observed improved power distribution as well as improved network scalability and operational lifespan.</p>