Dynamic Firewall Generation using Real -Time Analysis of Network Traffic with OC-SVM Algorithm

Authors

  • Elamaran E., Sudhakar B.

Abstract

With the number of internet connected devices increasing exponentially, the information about the activities of these devices and the services they use are obscure. It becomes vital to understand the network activity of each device to protect the network from security threats. Currently, the firewall used in home networks is static which performs filtering based on rules that were configured manually at the initial stage, while the threats posed to the network are dynamic in nature. There is a need for dynamic firewall systems that have the ability to adapt to the changes on the internet. The solution proposed in the paper involves analysis of real-time data on a server to formulate a set of dynamic rules which the routers will use. The use of real-time analysis allows for the rules to adapt towards the dynamic nature of threat posed by the internet. Several techniques have been developed to detect suspicious traffic such as, use of threshold, heuristics and statistical models. Although these techniques are easy to apply, they are not effective when used in dynamically changing environments. In this paper, we intend to capture real-time network data using custom firmware on the routers and classify it using One Class – Support Vector Machines (OC-SVM) model which is then used to generate dynamic configurations.

Published

2021-08-13

How to Cite

Elamaran E., Sudhakar B. (2021). Dynamic Firewall Generation using Real -Time Analysis of Network Traffic with OC-SVM Algorithm. Drugs and Cell Therapies in Hematology, 10(1), 1262–1269. Retrieved from http://www.dcth.org/index.php/journal/article/view/248

Issue

Section

Articles