Network security situation assessment based on BKA and cross dual-channel

Network security situation assessment based on BKA and cross dual-channel

연구 분야: Infrastructure

학회: The Journal of Supercomputing

Network security situation assessment (NSSA) has become increasingly critical due to the growing frequency and sophistication of network attacks. NSSA involves analyzing network threats and security incidents to support network administrators in decision-making and the implementation of protective strategies. To address the challenges of low assessment accuracy in current NSSA methods, we propose a novel model that integrates an enhanced black-winged kite algorithm (BKA) with a cross dual-channel framework. First, we develop a cross dual-channel architecture that combines a convolutional neural network with a bidirectional long short-term memory network. This structure effectively integrates temporal and spatial features; while, an attention mechanism highlights key information, thereby improving the accuracy of traffic classification. Second, the improved BKA is employed to optimize network parameters, further enhancing the model’s overall performance. Finally, the situation value is derived from the classification results and mapped to corresponding network security situation levels, completing the NSSA process. Experimental results on the NSL-KDD dataset demonstrate that the proposed model achieves notable improvements, with an accuracy of 83.66%, a recall of 80.04%, and an F1-score of 83.13%. Moreover, the proposed assessment method offers a more robust and comprehensive evaluation of the network’s overall security status, highlighting its potential for practical application.