A Decision Tree-Based Middle Ware Platform for Deploying Fog Computing Services

Abstract

Cloud computing has become a cost-effective and reliable distributed computing model for the end users to share IT resources from a pool of computational resources based on their demand in real time. Cloud computing offers advantages of rapid provisioning and release of resources with minimal effort. While cloud computing offers advantages of cost-effective access to sophisticated hardware and software, the turn around time is a hindrance because of unreliable physical layer connectivity especially for business located in a remote location. Not all business solutions need cloud services to the same extent and at all times as the local IT resources available may be sufficient to handle a business issue such as performing analytics on the locally stored data. Fog computing and Grid computing helps in achieving this. Fog computing aims to bring computational power to the edge of the network and by doing so reduces the operational cost and execution time at the cost of accuracy of results. However, a middle ware is required to determine whether an information query needs to be executed in the cloud or if it can be executed on a group of computers that are geographically closer to the business. In this paper, we propose and develop an intelligent middle ware platform that is based on decision trees to optimization the execution of any information query.