Review · Sensors · December 17, 2022

Research of Machine Learning Algorithms for the Development of Intrusion Detection Systems in 5G Mobile Networks and Beyond

Azamat Imanbayev, Sakhybay Tynymbayev, Роман Сергійович Одарченко, Sergiy O. Gnatyuk, Rat Berdibayev, Алимжан Байкенов

Abstract

The introduction of fifth generation mobile networks is underway all over the world which makes many people think about the security of the network from any hacking. Over the past few years, researchers from around the world have raised this issue intensively as new technologies seek to integrate into many areas of business and human infrastructure. This paper proposes to implement an IDS (Intrusion Detection System) machine learning approach into the 5G core architecture to serve as part of the security architecture. This paper gives a brief overview of intrusion detection datasets and compares machine learning and deep learning algorithms for intrusion detection. The models are built on the basis of two network data CICIDS2017 and CSE-CIC-IDS-2018. After testing, the ML and DL models are compared to find the best fit with a high level of accuracy. Gradient Boost emerged as the top method when we compared the best results based on metrics, displaying 99.3% for a secure dataset and 96.4% for attacks on the test set.

Network Security and Intrusion DetectionNetwork Packet Processing and OptimizationArtificial Immune Systems Applicationsintrusion detection systems5Gnetwork security

Research of Machine Learning Algorithms for the Development of Intrusion Detection Systems in 5G Mobile Networks and Beyond

Abstract

More by Azamat Imanbayev