Internet of Things (IoT) is a technique to connect each and everything to the Internet, and it provides anything at any time regardless of irrespective of the location of the thing. As we are moving towards the IoT, the number of sensors deployed around the world is also growing at a rapid pace and an accelerated growth is expected in future. An increasing number of IoT middleware solutions are being developed in both research as well as commercial environments. However, sensor search and selection still remains a critical requirement and a challenge owing to its diversity and heterogeneity. This text presents a new architecture for fog computing along with mathematical modeling and logical design to achieve the minimized latency. Various methods have been introduced so as to carefully choose the fog node that helps in reducing the overall delay and generating quick responses. Different methods have been proposed to carry out a low latency and energy efficient sensor selection process for IoT services in the fog environment associated with any application presented in the form of flow based process(FBP).