Journal
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
Volume 12, Issue 11, Pages 3337-3349Publisher
SPRINGER HEIDELBERG
DOI: 10.1007/s13042-021-01323-7
Keywords
DDoS; Deep learning; 5G; Intrusion detection; IoT security; Kalman backpropagation; Neural networks
Categories
Funding
- Hussein Bin Talal University, Jordan [2019/197]
Ask authors/readers for more resources
The high-speed data transmission of 5G wireless communication systems significantly impacts the performance of IoT networks. Emerging cybersecurity threats pose risks to data transfer and security. Utilizing neural networks for DDoS intrusion detection can help address these issues.
The fifth-generation (5G) wireless communication systems associating with the high achievable data-transfer speeds will significantly affect the performance of IoT networks. On one hand, the internet goes through a dramatic transaction period that shapes every aspect of our lives, industry, and business where cloud computing, smart cities, and the Internet of Things (IoT) play a significant role in the advancement of data transfer, storing, and processing. On the other hand, it plays a significant role in emerging advanced versions of different types of cybersecurity attacks especially that are novel, hard-to-detect, and that of distributive never cease-fire characteristics. To mitigate these concerns, we present a distributed denial-of-service (DDoS) intrusion detection model that can be implemented in IoT dynamic environments, providing an intelligent intrusion detection mechanism against the second biggest threat to data traffic and transfer on IoT networks. Kalman backpropagation neural network-based DDoS intrusion detection is proposed in this work. The framework is validated through various simulations via the most up to date CICDDoS2019 dataset to demonstrate the effectiveness of the solution in terms of intrusion detection. the proposed solution achieved an average detection accuracy of 94% with 0.0952 false alarm rate and 97.49%, 91.22% for detection rate, and precision respectively.
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