Advanced Big Data Security Analytics Using Deep Learning and Adaptive Cyber Defense Mechanisms

Abstract

Growing deployment of cloud computing, Internet of Things (IoT) networks, and volume and complexity of large-scale digital infrastructures have greatly augmented the number and types of cybersecurity threats encountered. The proliferation of cloud computing, IoT networks, and large-scale digital infrastructure has greatly increased both the quantity and complexity of cybersecurity threats faced. Traditional security monitoring systems have a lot of trouble managing large datasets and being able to recognize advanced attack patterns in changing environments because the data is heterogeneous and size is relatively large. To solve the above problems, the Dynamic Threat Propagation Suppression Framework (DTPSF) is proposed in this paper. The proposed DTPSF combines large volume security data analysis, deep learning-based security intelligence, multi-model classification and adaptive cyber defence mechanisms, within a single system. First, the security information is gathered from different data sources and preprocessed and presented in a feature representation. Next, deep learning analytics filters the behavior from the security events and Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and Hybrid Deep Neural Network (Hybrid DNN) classifiers are employed for intelligent threat classification. The framework also covers a threat severity evaluation mechanism that helps identify potentially malicious activities and automatically create a suitable defense response. Unlike attacks detection approaches which are predominantly utilized, DTPSF is the continuous security awareness and adaptative mitigation process depending on the developing attacks condition. It is envisioned for providing capability at scale in managing cybersecurity in today's big data setting, and for improving the sophistication and agility of cybersecurity analysis and defense. This is why DTPSF has developed a holistic solution for enhancing security analytics and cyber defence in the intricate and information overload digital environments. The suggested DTPSF attained an overall accuracy of 98.47%.