CLOUD COMPUTING INFRASTRUCTURE DATA SECURITY: AN OPTIMIZED VIRTUALIZATION MODEL

The study suggests using an improved and robust virtualization model to apply an intelligent-based design to data security monitoring in a cloud computing infrastructure. The primary problem with cloud computing, the concurrently concerning malicious activity, makes this technology notable. When we discuss an architecture that allows for easy, on-demand network access to a shared pool of reconfigurable computing resources—such as servers, networks, storage, apps, and services—we are referring to cloud computing. This requires little administration work or communication between the service provider and customer to quickly deploy and discharge. A virtualized approach was created in this work to improve the monitoring of data security. Adopting the Structured System Analysis and Design Methodology, dataflow diagrams, use-case diagrams, sequence diagrams, and diagrams created with the Unified Modelling Language (UML) were utilized to accomplish the desired design. Robust repositories provided five hundred (500) datasets, of which thirty percent were used for training and seventy percent were used for testing. The number of adopted technologies, the number of adopted design tools, the number of adopted algorithms, and the number of tested records were all used as parameters to analyze and assess the performance of both systems. According to the performance review, the new system performed better than the old one, achieving an accuracy rate of 1.07% as opposed to the old system's 0.48% accuracy rate. The recently created model focused specifically on financial fraud and was intended to detect fake data in cloud computing infrastructure. Because financial frauds against property involve the illegal transfer of property ownership for an individual's personal use and benefit, this study could be helpful to corporate organizations, anti-corruption agencies, and researchers who have a keen interest in the subject matter. Additionally, the new system was further optimized with the use of deep neural networks and logistic regression techniques.