| Autor: |
Kumar, Prahlad, Bagga, Harnoor, Netam, Bhuneshwar Singh, Uduthalapally, Venkanna |
| Předmět: |
|
| Zdroj: |
Wireless Personal Communications; Jan2022, Vol. 122 Issue 1, p87-108, 22p |
| Abstrakt: |
Internet of Things is one of the most versatile technologies in existence today. It has taken over our day to day activities and thus has many applications that are designed to make life easier and simpler. Partly because IoT is new, it is replete with insecurities and vulnerabilities. Due to the lack of fundamental security controls, and the integration of real-world objects with the Internet, IoT devices are facile targets for cyber-criminals and other aggressors. This means that these vulnerabilities can be exploited for hacking, adding to Botnets, and then used to launch DoS and DDoS against organizations. To provide security from DoS and DDoS attacks, various solutions have been proposed. In this paper, Machine Learning, as well as Deep Learning algorithms, have been employed to analyze the DoS and DDoS attacks. The Bot-IoT dataset of the Centre of UNSW Canberra Cyber was used for training purposes. ARGUS software was used to generate the features from the pcap files of UNSW. A testbed was setup using 20 devices and generated dataset. From the result, the best accuracy of attack classification is 99.5% and 99.9% for Deep Learning and Machine Learning algorithms respectively. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
Full text from SpringerLink