Application of Artificial Intelligence for DDoS Attack Detection and Prevention on Cyber Physical Systems Using Deep Learning

Abstract

There are new discoveries everyday as technology is improved. As these improvements are used in solving one particular issue, more problems are then discovered. The adoption of cyber physical systems (CPSs) creates scenarios of cyberattacks, particularly distributed denial of service (DDoS) attacks. Sensing the attacks on CPSs becomes more challenging due to the incessant cyberattacks. Hence, artificial intelligence (AI) is being used for digital transformation to mitigate these attacks. AI technique was implemented to prevent distributed denial of service (DDoS) attacks using deep learning (DL) algorithms, such as convolutional neural network (CNN), long short-term memory (LSTM), and dense and gated recurrent units (GRU). DL models were used to monitor and potentially flag incoming attacks. Hold-out technique with ratio 80 to 20 was used for performance evaluation using accuracy, loss function, precision, recall, and root mean square error (RMSE) metrics. DL performance results showed the best results with 99.92% accuracy, 0.0037 loss function, and 0.026 RMSE for classification by LSTM in the training phase, whereas in the testing phase the results were 99.92% accuracy, 0.0058 loss function, and 0.0278 RMSE. LSTM outperformed other DL algorithms for the mitigation of DDoS attacks. Consequently, the developed model will help secure cyber space, particularly against DDoS attacks.