SDN-Driven Traffic-Aware Load Balancing for Latency Optimization in WiFi-Enabled Smart Cities

The rise of wireless smart gadgets and constant connectivity boosts smart cities. WiFi remains the most widely used wireless technology for IoT integration in smart cities, driven by its pervasive access points and minimal deployment expenses. Smart city applications offer a wide range of services, each with distinct quality-of-service (QoS) demands. This study focuses on packet delivery latency as a key quality of service (QoS) metric, which has a major impact on various time-sensitive smart city applications. To address this, the paper proposes leveraging Software-Defined Networking (SDN) to manage the traffic distribution across WiFi access points (APs), maintaining balanced load conditions within a city-wide network of WiFi-connected Internet of Things (IoT) gateways or fog nodes. These gateways collect data from IoT or smart city devices over wireless links and relay the packets through the municipal WiFi infrastructure to backend servers or control centers. The developers designed three SDN-based algorithms to reduce packet forwarding delays and maintain balanced traffic distribution across access points (APs). Researchers develop and evaluate the algorithms in a simulated environment using Anaconda Jupyter Notebook, emulating WiFi communication without imposing any additional requirements on IoT gateways (Wi-Fi clients) or access points (APs). This means no need for support of specific roaming protocols or bandwidth-intensive signaling, such as probe packet exchanges. Extensive software-based experiments show that the SDN controller, using the proposed algorithms, reduces packet forwarding latency at IoT gateways by detecting the gateway experiencing the highest delay and rerouting traffic to the least-loaded available access point (AP) without interruption. Results demonstrate significant improvements in latency reduction compared to traditional non-SDN or static AP selection methods.