PERFORMANCE ANALYSIS OF VGGNET FOR DOS AND DDOS DETECTION IN WIRELESS SENSOR NETWORKS

Abstract

Intrusion detection have become absolutely essential in today’s IT infrastructure, primarily, because they play a critical role in maintaining network security, these systems offers efficient and high-performance solutions for spotting and mitigating a broad spectrum of network attacks. As communication technologies advance rapidly and become more widespread, the risk of malicious users exploiting network vulnerabilities has grown substantially. This surge in potential threats has paved the way for highly sophisticated, automated, and distributed attacks, which traditional IDS frameworks find challenging to counter effectively. Given these advanced threats, there is an urgent need to explore and implement more robust and adaptive IDS methodologies. This contribution examines and assesses the VGGNet deep learning model in detail, with a focus on intrusion detection tasks. We concentrate on strengthening the security of wireless sensor networks (WSNs), which are vulnerable to denial-of-service (DoS) and distributed denial-of-service (DDoS) attacks because of their low processing power and frequent deployment in hostile situations. We use our suggested approach to two well-known datasets, CICDDoS-2019 and WSN-DS, to assess its performance. The WSN-DS dataset, which captures a range of legitimate and malicious traffic patterns, was created especially for WSN situations. Meanwhile, a thorough collection of DDoS attack scenarios is offered by the CICDDoS-2019 dataset. We use these datasets to undertake a thorough performance assessment of our VGGNet-based model, including common measures like recall, precision, accuracy, F1-score, and confusion matrix. The results of our tests demonstrate that the VGGNet model performs remarkably well in identifying intrusions, obtaining an accuracy rate of 0.979. This high degree of accuracy demonstrates how deep learning models, such as VGGNet, have the potential to significantly improve the resilience and dependability of IDS in defending contemporary network infrastructures against changing cyberthreats.