Detecting HTTP Vulnerabilities in IoT-based Precision Farming Connected with Cloud Environment using Artificial Intelligence

Abstract

Nowadays there is significant growth in the Internet of Things (IoT) enabled applications in connection with cloud computing is a framework that enables IoT users to access computing resources on-demand. IoT provides a new dimension in the area of the smart farming domain from the last decade. In such precision farming applications, the on-field sensor node collects the periodic data and transmits it to the remote station over the wireless channels. Data collected at the wireless station is huge and hence they are stored into the cloud servers. Further in the cloud computing environment, various data processing related operations performed based on dynamic demands for the decisionmaking process. The cloud computing communications established using the HTTP protocol and hence various HTTP vulnerabilities introduced to harm farming data. There are various methods are introduced for the detection of such attacks with drawbacks less accuracy rate and higher detection errors. Thus applying such traditional methods to detect the attack is nightmare and required efficient approach for detection in IoT connected cloud computing domain. This paper proposed the detecting HTTP vulnerabilities in a Cloud computing based on Artificial Intelligence (AI). The proposed system comprises the following phases: data gathering, data cleaning, feature selection and computation, and classification. In the data cleaning stage, the acquired cloud traffic is cleaned to remove the unwanted noisy data. In the features extraction phase, the key features selected and normalized using the statistical methods. In the classification, stage data is grouped into normal and attack using the J48 Decision tree classifier. The experimental result shows the proposed approach achieved superior performance over the previous methods.