Abstract
With the significant growth of the cyber environment over recent years, defensive mechanisms against adversaries have become an important step in maintaining online safety. The adaptive defense mechanism is an evolving approach that, when combined with nature-inspired algorithms, allows users to effectively run a series of artificial intelligence-driven tests on their customized networks to detect normal and under attack behavior of the nodes or machines attached to the network. This includes a detailed analysis of the difference in the throughput, end-to-end delay, and packet delivery ratio of the nodes before and after an attack. In this paper, we compare the behavior and fitness of the nodes when nodes under a simulated attack are altered, aiding several nature-inspired cyber security-based adaptive defense mechanism approaches and achieving clear experimental results. The simulation results show the effectiveness of the fitness of the nodes and their differences through a specially crafted metric value defined using the network performance statistics and the actual throughput difference of the attacked node before and after the attack.
Subject
Information Systems and Management,Computer Networks and Communications,Modeling and Simulation,Control and Systems Engineering,Software
Reference25 articles.
1. Machine Learning Techniques for Anomaly Detection An Overview;Omar;Int. J. Comput. Appl.,2013
2. A comparative evaluation of outlier detection algorithms: Experiments and analyses;Domingues;Pattern Recognit.,2017
3. Hodge, V., and Austin, J. (2018). An Evaluation of Classification and Outlier Detection Algorithms. arXiv.
4. Outlier Detection Based on Gaussian Process with Application to Industrial Processes;Wang;Appl. Soft Comput.,2019
5. Performance evaluation of unsupervised techniques in cyber-attack anomaly detection;Meira;J. Ambient. Intell. Humaniz. Comput.,2020
Cited by
7 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献