Detection of cyber-attacks and node recovery in mobile networks.
Abstract
According to the current statistics, 83.96% of the world’s population owns a smartphone. The number raises daily due to the increasing need for internetpowered services. For various reasons including monetary reasons, and market competition, malicious actors attack both end user devices and network infrastructures to disrupt communication channels.
In this project, we are focusing on an infrastructure targeted Denial of Service attack known as Signaling amplification attack. The network attach procedure involves a large collection of data between user equipment, radio access network and mobility management entity. Cybercriminals have ability to initiate this same process with the intention of overwhelming the network infrastructure hence denying service to the legit network users.
This report therefore presents the development and simulation of a deployed machine learning model that will enable timely detection of a signaling amplification attack, isolation of the malicious source from the network and recovery mechanism when the particular node’s behaviour normalizes.
An intrusion detection machine learning model is trained with KDD99 Dataset.