首页|Development of hybrid weighted networks of RNN and DBN for facilitating the secure information system in cyber security using meta-heuristic improvement
Development of hybrid weighted networks of RNN and DBN for facilitating the secure information system in cyber security using meta-heuristic improvement
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
Springer Nature
Abstract As communication and information technologies are integral to everyone’s daily activities, the significance of cybersecurity has become more pronounced due to the growing vulnerability of these technologies to cyber threats. Traditional cyber security systems use various preventive measures to secure the information and trust authentication methods are used to provide the essential security measure against cyber attacks. These methods are efficient and are also equipped to perform in real-world scenarios. However, the conventional cyber security system does not provide essential security against all types of cyber attacks as they are nature-distributed for controlling the systems. Securing these Distributed Control Systems is highly significant for providing a secure and risk-free operation of the connected systems from cyber attacks and other threats. Therefore, a novel method of risk prediction and risk mitigation is developed using the heuristic-based Hybrid Deep Weighted Networks for protecting the data in the information system. The recommended work is based on risk analysis and a cyber security framework built around the information technology security system. This model aimed to design the cyber security system by mitigating all the threats in that particular information system. The main aim is to predict the risk level and mitigate the security threats completely from the system. To achieve this, initially, the data are gathered from different sources and given to the HDWN. The HDWN is developed by combining the Deep Belief Network and a Recurrent Neural Network. These two networks help to predict the risk values of the threat. To attain the enhanced results, the parameters in this model are optimized by using a hybrid algorithm known as African Vultures with Water Wave Optimization, which is developed by combining the Water Wave Optimization algorithm with the African Vulture Optimization Algorithm. Another intention of this model is to mitigate the threat present in the system. Based on the predicted risk value, the system generates warning signals to alert the admin to block the communication. Thus, the threat and risk from the system are predicted and mitigated without interrupting the system’s performance. Finally, the performance validation is performed on the developed model by comparing it with diverse approaches, and the results demonstrate that the proposed model provides impressive outcomes in ensuring data security.
NETL
NSTL
Springer Nature
