An Evaluation of Transformer Models for Early Intrusion Detection in Cloud Continuum

With the increasing popularity of the cloud continuum, the security of different layers and nodes involved has become more relevant than ever. Intrusion detection systems, are one of the main tools to identify and intercept intrusion attacks. Furthermore, identifying the attacks in time, before they...

Full description

Saved in:
Bibliographic Details
Main Authors: Islam, Md Mahbub, Ahmad, Tanwir, Truscan, Dragos
Format: Conference Proceeding
Language:English
Subjects:
Cloud computing
Computational modeling
Computer architecture
Computers
deep learning
Intrusion detection
Intrusion detection systems
monitoring
Roads
transformer-architecture
Transformers
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:With the increasing popularity of the cloud continuum, the security of different layers and nodes involved has become more relevant than ever. Intrusion detection systems, are one of the main tools to identify and intercept intrusion attacks. Furthermore, identifying the attacks in time, before they are completed, is necessary in order to deploy countermeasures in time and to limit the losses. In this work, we evaluate the use of transformer models for implementing early-detection signature-based detection systems targeted at Cloud Continuum. We implement the approach in the context of our tool for early detection of network intrusions and we evaluate it using the CICIDS2017 dataset and MQTT-IDS-2020. The results show that transformer models are a viable alternative for early-detection systems and this will pave the road for further research on the topic.
ISSN:2380-8004
DOI:10.1109/CloudCom59040.2023.00052