Journal
IEEE ACCESS
Volume 8, Issue -, Pages 126215-126227Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/ACCESS.2020.2995820
Keywords
Computer crime; IP networks; Support vector machines; Real-time systems; Probabilistic logic; Internet; DDoS; network attacks; IP networks; security; dataset
Categories
Funding
- Alan Turing Institute [EP/N510129/1]
- PETRAS National Centre of Excellence for IoT Systems Cybersecurity [EP/S035362/1]
- EPSRC [EP/R007195/1] Funding Source: UKRI
Ask authors/readers for more resources
Computer networks are subject to an unprecedented number and variety of attack, the majority of which are distributed denial of service (DDoS). The nature and mechanisms employed in these DDoS attacks continually change, creating a significant challenge for detection and management. To address this evolving nature of attacks, approaches are required that can effectively detect and mitigate emerging attacks. In this paper, we provide a mechanism that not only detects the presence of a DDoS attacks but also identifies the route of attack and commences a process of mitigation at the initial stage of identification. The proposed research involves an optimized SVM classification algorithm integrated with SNORT IPS to provide prevention mechanisms for the entire network when subject to DDoS attack. The proposed IPS method allows traffic identified as legitimate to pass through the network, whereas suspect traffic is flagged and has to go through an identification system. We present the algorithm with experimental results that show better performance than simple Snort IPS, Probabilistic Neural Network (PNN), Back Propagation (BP), Chi-square, and PSO-SVM in terms of accuracy, exposure and specificity. These results show that the average accuracy rate of our method is 97 percent.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available