Internets of Things (IoT) are network of interconnected devices and communicate through internet. IoT constitutes huge number of heterogeneous devices and is prone to many issues. Devices in IoT senses enormous information that needs lots of network resources. In this work we have designed a Fuzzy Interference System (FIS) for classification of the jobs in to emergent and non-emergent jobs at local processor. The job values are sent to the global processor for requesting the resources of the cloud. Global Processor estimates the resource requirement and available resources of the cloud in terms of communication and computational cost. Based on the type of the job Global processor allocates network resources by using communication and computational cost. In order to evaluate the performance of the proposed system, it is simulated using C++.