Accurate Information Analysis Scheme for High Level Security Provisioning in Cloud Environment

Many experts agree that cloud computing represents a major technological advance for the Internet and related computer systems. This shift has led many companies to outsource both their information and computer needs. Such systems need to be secure in terms of integrity, availability, and confidentiality, the three cornerstones of information security. Cloud computing gives users access to virtualized, services that are scalable that are adaptable and need no upkeep. It's conceivable that the protocols and infrastructure supporting cloud computing are inadequate. The vulnerabilities are being exploited by attackers. There has been much discussion about how vulnerable cloud infrastructure is to distributed denial-of-service (DDOS) attacks. A denial-of-service (DDOS) attack is characterised by the simultaneous transmission of a large number of information packets. Because all of the network's and servers' resources have been exhausted, an denial of service has occurred. In an attempt to identify DDOS assaults early on, a number of methods have been proposed and analysed. It's possible that an attack detection system need only use a single machine learning classification method, however utilising many models is recommended. In this study, we suggest using several neural network classifiers together. The proposed technique uses a Hybrid Mechanism for intrusion detection, which combines the Teacher Learning Optimisation (TLO) using a Fully Recurrent Neural Network (FRNN). We looked at information from the CICDDOS2019 information set, which included a total of 32,000 incidents (8,450 benign and 23,550 DDOS attack occurrences). Experiment results showed a 98.02 percentage of accurateness, 97.5 percent sensitivity, and 98.5 percent specificity.