The DDoS Attack Detection for Information Systems of High-Penetration Renewable Energy Grids Based on TCN-BiLSTM

Abstract In high-penetration renewable energy grids, while providing efficient and intelligent energy supply and consumption management, they also face significant cybersecurity threats, particularly Distributed Denial of Service (DDoS) attacks. Due to the high dependence of the grid’s information systems on communication technologies, these attacks pose substantial risks to the stability and reliability of the grid. To tackle this issue, this paper introduces a novel approach for detecting DDoS attacks, leveraging Temporal Convolutional Networks (TCN) in combination with Bidirectional Long Short-Term Memory (BiLSTM). This method is particularly tailored for the detection of DDoS attacks within information systems of high-penetration renewable energy grids, where the complexity of traffic patterns requires advanced, efficient solutions. First, an Ant Colony Optimization (ACO) algorithm combined with Dynamic Pheromone Adjustment and Incentive Mechanism (DIAACO) is used for feature selection, extracting the most representative features from the raw data while eliminating redundant and irrelevant parts. Next, the TCN module is employed to model the data, automatically extracting key features and effectively handling the complexity of the grid data. Subsequently, the BiLSTM module is applied to further model the features extracted by TCN, capturing long-term dependencies within the data and significantly improving the DDoS attack detection capability and robustness in dynamic and complex network environments. Finally, the output from the BiLSTM is used for DDoS attack detection in high-penetration renewable energy grids. Experimental results based on the CIC-DDoS2019 public dataset show that the proposed method achieves accuracy, precision, recall, and F1 scores of 99.89%, 99.86%, 99.88%, and 99.87%, respectively, confirming the effectiveness of the proposed method in DDoS attack detection for high-penetration renewable energy grid information systems.