Enhancing DNS Light Attack Detection through RNN-RF Integration for Imbalanced Data

In order to maintain the network’s security and ensure that users continue to have faith in it, it is of utmost importance to protect it against DNS lightweight attacks. The application of deep learning is a helpful method that may be utilized to strengthen defense against threats of this nature. Through the utilization of deep learning techniques, networks are able to more effectively detect and eliminate DNS lightweight threats. To prevent anyone from stealing data and gaining access without authorization, these methods search for patterns in DNS traffic to identify inappropriate conduct. Deep learning has the potential to considerably simplify the process of locating and preventing DNS lightweight attacks, which would result in networks that are both more secure and more dependable. Using deep learning to enhance the safety of networks is one approach that may be used to accomplish this goal. With Recurrent Neural Networks (RNN) and Random Forest working together, DNS lightweight attacks are much less likely to succeed. This strategy effectively identified and addressed these potential risks by taking into consideration a variety of criteria, including loss, accuracy, val-loss, and valaccuracy. When these technologies are combined, network security improves because DNS traffic trends may be analyzed to detect signals of malicious conduct. When these variables are utilized, the model becomes more robust and better able to detect and prevent DNS lightweight attacks. This improves network security overall. The research effort utilized the Bell-DNS-EXF-2021 dataset to gain a comprehensive understanding of DNS network threats and to devise strategies to mitigate them.