BNID: A Behavior-based Network Intrusion Detection at Network-Layer in Cloud Environment

Abstract

Security has become one of the crucial issues in today’s new technological environment such as cloud computing. In recent years, research work has been done to tackle various cloud security issues. This paper proposes a light weighted and adaptable intrusion detection approach named as Behavior-based Network Intrusion Detection (BNID) at network-layer in cloud.The behavior analysis of traffic is performed at Cloud Network Node (CNN) to detect the intrusions. A security framework is proposed for deployment of BNID in cloud. The need of placement of IDS in each and every tenant virtual machine (TVM) is eliminated. BNID uses statistical learning techniques with feature selection for traffic behavior analysis and does not require the extensive monitoring of memory writes. Information Technology Operations Center (ITOC) attack dataset is used to validate our approach. BNID achieves an accuracy of 98.88% with 1.57% false positives which seems to be promising.