The execution of smart cities around the globe has risen due to the steady connectivity and increased number of wireless devices. Due to low-cost network construction and simple technical implementation, Wi-Fi networks have become a dominant wireless technology to enable the connectivity of Internet-of-Things (IoT) in smart cities. There are a number of services and applications running in smart cities with different demands of the quality of service (QoS). The paper focuses to address the latency problem, which is a key performance metric regarding QoS, in time sensitive applications in smart cities. The emerging paradigm, Software-defined Networking (SDN) is extended for Wi-Fi networks to ensure fairness of traffic load among the access points (AP). We propose three algorithms based on service time, M/G/1 analysis and AP selection to determine the packet transmission delay, packet latency rates and choosing a least loaded destination AP respectively. The optimization of load among the APs ensures a reduced packet latency factor, when a communication link is formed between the smart city IoT devices and the APs. A symmetric load index and a reduced packet latency rate is maintained between the IoT devices and the OpenFlow enabled APs using three software-defined algorithms designed in this study. A Linux based software-defined testbed is developed to ensure the credibility of the algorithms developed. Extensive experimentation using the hardware devices confirm that the proposed algorithms are efficient enough to reduce the latency rate and enhance the throughput rate by 17%, 13% and 9% when compared to received signal strength indicator scheme (RSSI), Po-Fi scheme and aggregated Wi-Fi scheme respectively, by shifting the wireless traffic load from a higher packet latency IoT device to a least loaded AP.