International Journal of Knowledge Based Computer System

Anshu Mali Bhushan ¹, Harsh Kumar ², Devendra Singh Chauhan ³

Affiliations
1 Research Scholar, Computer Application Department, Himalayan Garhwal University, Pauri Garhwal, Uttarakhand, IN
2 Faculty of Computer Application, Himalayan Garhwal University, Pauri Garhwal, Uttarakhand, IN
3 Assistant Professor, Department of Physics, Dr. PDBH Government PG College, Kotdwar, Uttarakhand, IN

Abstract

The detection part is called the Hypervisor Scanner, which is programmed to detect malicious insiders. Using the feed forward neural network, the hypervisor scanner is generated and trained with a supervised learning algorithm referred to as the Levenberg-Marquardt algorithm. The three criteria are considered for service level agreement, such as bandwidth requirement, memory consumption and storage space. The Hypervisor Scanner can detect malicious insiders in cloud systems that breach SLA and suffer from an insecure cloud administrative domain that lacks control over the cloud service provider (CSP). Here, the Hypervisor Scanner is constructed using the biologically inspired classification approach referred to as artificial neural network modelling. ANN teaching uses the Levenberg-Marquardt learning algorithm. The LM algorithm works by minimising the average square error to boost the detection system. For detecting malicious insiders, the Hypervisor Scanner uses threshold values for SLA parameters. It is known from performance review and comparison that the Hypervisor Scanner is the appropriate one with high detection accuracy and low false alarm rate for the detection of malicious insiders. This can therefore be effective, robust and realistic in the detection of malicious insiders in and around the cloud world.

Keywords

Architecture, Cloud Computing, Commercial and Technological, Hypervisor Scanner

Full Text