With the explosive growth of Internet of Things (IoT) using WiFi networks along with their huge data flows (especially Bigdata using TCP connections), the significant challenges are the application performance and network security. Bigdata comes in form of varying volume, velocity, etc. and is very challenging to manage with traditional networks. Therefore, we advocate Software-defined networking (SDN) paradigm in this paper. Using SDN, firstly, from security perspective, we are able to diagnose Bigdata TCP streams that may come from both attack or non-attack sources. Secondly, when the Bigdata TCP streams come from legitimate sources, SDN can help in maintaining Quality of Service (QoS) to particular flow or application. In this paper, we have proposed a Policy-based framework that maintains the security as well the flow specific QoS requirement in SDN enabled IoT network. In our network settings, we proposed an algorithm at WiFi Access Point (AP) or at network edge router, to learn the incoming traffic from different Things and then takes appropriate action/s based on the policies in place. A mathematical model is developed considering TCP CUBIC streams over WiFi networks to understand and evaluate our idea. Our extensive simulation results demonstrate how we jointly enhance the security and effectively maintain the desired QoS of the streams in real time.