Network intrusion detection and prevention strategy with data encryption using hybrid detection classifier

Advances in technology and communication have led to an increase in data and network sizes. As a result, many intrusions (attacks) are created and pose more challenges to network security. In order to protect the network's confidentiality, integrity, and availability from potential intrusions, an Intrusion Detection System (IDS) was employed to examine network traffic. Despite significant research efforts, IDS still has difficulties detecting novel intrusions and improving detection accuracy while lowering false alarm rates. To solve these problems, a novel intrusion detection approach with data encryption is proposed to detect intrusion and protect data from attackers. The operational process of the proposed approach is divided into three phases. First phase: Intrusion identification from the network through a novel hybrid machine learning and deep learning approach. Second phase: an advanced deep learning approach is utilized to detect the types of attack. The attack types that occur in the network are predicted, and an automatic alarm message is sent to the users. Third phase: advanced encryption model is utilized to store the data securely. The users from the network could able to access data from the storage after complete the four barrier authentication process. Each phase is separately analyzed and compared to existing approaches. Proposed hybrid intrusion detection phase provide 0.97 accuracy and 0.02 false positive rate, as well as proposed attack type detection phase offer 0.96 accuracy and 0.01 false positive rate. Observed values prove that the proposed model is most suitable for all network intrusion detection and data encryption.