QoS-aware routing in software-defined internet of things: A comparative evaluation of Dijkstra and bellman–ford under heterogeneous traffic

Abstract

The rapid growth of Internet of Things (IoT) infrastructures has created heterogeneous communication environments where multiple applications share network resources, making reliable Quality of Service (QoS) increasingly challenging, especially when routing must support diverse traffic requirements. This study evaluates the performance of two shortest-path routing algorithms—Dijkstra and Bellman-Ford—within a Software-Defined Internet of Things (SD-IoT) architecture to analyze their impact on QoS and resource utilization under heterogeneous traffic. An experimental SD-IoT testbed was implemented using the Ryu controller and Mininet emulator, where each algorithm was integrated and tested separately on a programmable topology. Three IoT applications—e-health, intelligent transportation, and industrial control—were generated simultaneously. The evaluation considered latency, packet loss, link load distribution, and CPU/RAM usage at controller and switch levels. Results show distinct behaviors: Dijkstra achieves lower latency and packet loss due to faster path computation, while Bellman-Ford introduces slightly higher delays and resource consumption because of its iterative process, yet remains stable under varying loads. Both improve link load distribution compared to default routing. Overall, routing efficiency depends on convergence behavior and traffic heterogeneity; Dijkstra suits latency-sensitive applications, whereas Bellman-Ford is more appropriate for adaptive routing scenarios, contributing to QoS-aware SD-IoT design.