AN EFFICIENT HYBRID MODEL FOR DETECTING DISTRIBUTED DENIAL OF SERVICE (DDOS) ATTACKS IN CLOUD COMPUTING USING MULTIVARIATE CORRELATION AND DATA MINING CLUSTERING TECHNIQUES
Issues in Information Systems · 2018
The distributed nature of cloud computing makes it vulnerable and prone to sophisticated distributed intrusion attacks such as Distributed Denial of Service (DDoS) attacks.In order to detect those network attacks and respond swiftly, there must be a reliable defense system designed to distinguish anomalies embedded in legitimate traffic.In order to ensure the high availability of any offered services, the data center resources must be protected from DDoS threats.The existing solutions for monitoring incoming traffic and detecting DDoS attacks have excessive false alarms and become ineffective in early detection of high level flooding attacks and resolution of the cloud service availability issues.Therefore, it is necessary to devise a model that can detect DDoS attacks and serve the legitimate users with available resources with minimal downtime.This research paper addresses this need by investigating the multivariate correlation among the selected and ranked features.This study presents the promising performance results of our proposed comprehensive hybrid solution model using DBSCAN and Entropy, discusses the research findings, and visualizes the experimental results to show the degree of parametrical dependency among the selected features and the effectiveness of our multivariate correlational approach.