Anomaly Detection in the Cloud Environment with Clustering Optimization Model for Attack Detection in IDs

In recent decades the role of the Internet in day-to-day activities are rapidly increased due to the accessibility and completion of tasks with in fraction of a second, as well as the banking and billing process proceeds online. Due to the character of openness in networks, security is the big threat to users. The intruders enter into the network in different manners to get access or change the original contents as well as block the connection, etc. The intrusion detection (IDS) process is a mandate to overcome the issues, the IDS is classified as misuse detection as well as anomaly detection. This paper constructed a framework for anomaly detection in the cloud environment. The constructed framework model uses the Rule-based Bayesian Optimization (RBBO) for the attack and malicious activities detection. The RBBO model uses the weighted rule-based approach with the estimation of mean and standard deviation in the dataset attributed to the attack and anomaly detection. Based on the generated rules the conditions are provided within the Bayesian Network for the dimensionality reduction. Through implementation of the Bayesian Network classification of attacks are performed for the KDD cup and real-time dataset. The comparative analysis expressed that the RBBO model performance is effective for the detection of anomalies and attacks in the cloud with the reduced false positive rate.