Anomaly Detection System for Users and Devices Involved in Internet of Things Integrated Cloud Applications Using Machine Algorithms

Recently, due to their application nature, devices in the Internet of Things (IoT) networks also need to be involved in the cloud. IoT devices generally work on top of the Internet, whereas today, the Internet is called a cloud because it provides an uncountable number of services to the public. Software, Platform, Infrastructure, and Service are the base classes of cloud services that support IoT devices. It is well known that the security level of the cloud is deficient because anyone can access it from anywhere at any strategy. Like the cloud, IoT networks also permit any device to join anytime from anywhere. This vulnerability of IoT networks is a cause for concern, and various conventional and optimisation methods have proposed multiple algorithms for increasing the IoT Cloud's security level, which have been developed for particular applications. This problem is solved by thoroughly examining the IoT devices, user information, and meta-information to detect anomalies in the cloud applications, ensuring the validity and effectiveness of our proposed solution. The IoT network is analyzed using an NS2 simulator, and machine learning algorithms carry out the data analytics to select the best one that outperforms the others in anomaly detection. The machine learning algorithms were implemented in Python and experimented with it to check their efficiency.