Preliminary Results in Using Attention for Increasing Attack Identification Efficiency

Tanwir Ahmad; Dragos Truscan; Jüri Vain

doi:10.1109/ICSTW58534.2023.00038

Preliminary Results in Using Attention for Increasing Attack Identification Efficiency

Tanwir Ahmad, Dragos Truscan, Jüri Vain

Information Technology

Research output: Chapter in Book/Conference proceeding › Published conference proceeding › Scientific › peer-review

3 Citations (Scopus)

157 Downloads (Pure)

Abstract

In previous work, we proposed an end-to-end intrusion early detection system to identify network attacks in real-time before they are complete and could cause more damage to the system under attack. To implement the approach, we have used a deep neural network which was trained in a supervised manner to extract relevant features from raw network traffic in order to classify network flows into different types of attacks. In this work, we discuss the initial results of the benefits that an attention mechanism brings to the classification performance and the capacity of the network to detect attacks earlier. We empirically evaluate our approach on the CICIDS2017 dataset. Preliminary results show that the attention mechanism improves both the balanced accuracy of the classifier as well as the early detection of attacks.

Original language	English
Title of host publication	Proceedings - 2023 IEEE 16th International Conference on Software Testing, Verification and Validation Workshops, ICSTW 2023
Publisher	IEEE
Pages	159-164
Number of pages	6
ISBN (Electronic)	9798350333350
ISBN (Print)	979-8-3503-3336-7
DOIs	https://doi.org/10.1109/ICSTW58534.2023.00038
Publication status	Published - May 2023
MoE publication type	A4 Article in a conference publication
Event	IEEE International Conference on Software Testing, Verification and Validation - Dublin, Ireland Duration: 16 Apr 2023 → 20 Apr 2023 Conference number: 16 https://conf.researchr.org/home/icst-2023

Publication series

Name	IEEE International Conference on Software Testing, Verification and Validation Workshops
Publisher	IEEE
ISSN (Print)	2159-4848

Conference

Conference	IEEE International Conference on Software Testing, Verification and Validation
Abbreviated title	ICST
Country/Territory	Ireland
City	Dublin
Period	16/04/23 → 20/04/23
Internet address	https://conf.researchr.org/home/icst-2023

Funding

This work was made possible with funding from the European Union’s Horizon 2020 research and innovation programme, under grant agreement No. 957212 (VeriDevOps). The opinions expressed and arguments employed herein do not necessarily reflect the official views of the funding body. This work was made possible with funding from the European Union's Horizon 2020 research and innovation programme, under grant agreement No. 957212 (VeriDevOps). The opinions expressed and arguments employed herein do not necessarily reflect the official views of the funding body.

Keywords

Attention Mechanism
Deep learning
Convolution Neural Network
Early Intrusion Detection System

Access to Document

10.1109/ICSTW58534.2023.00038

EarlyAttentionAccepted author manuscript, 267 KBLicence: Publisher rights policy

https://urn.fi/URN:NBN:fi-fe2023070582968

Cite this

Ahmad, T., Truscan, D., & Vain, J. (2023). Preliminary Results in Using Attention for Increasing Attack Identification Efficiency. In Proceedings - 2023 IEEE 16th International Conference on Software Testing, Verification and Validation Workshops, ICSTW 2023 (pp. 159-164). (IEEE International Conference on Software Testing, Verification and Validation Workshops). IEEE. https://doi.org/10.1109/ICSTW58534.2023.00038

Ahmad, Tanwir ; Truscan, Dragos ; Vain, Jüri. / Preliminary Results in Using Attention for Increasing Attack Identification Efficiency. Proceedings - 2023 IEEE 16th International Conference on Software Testing, Verification and Validation Workshops, ICSTW 2023. IEEE, 2023. pp. 159-164 (IEEE International Conference on Software Testing, Verification and Validation Workshops).

@inproceedings{26620afeda9b4c3b982a2dc1ecd4014d,

title = "Preliminary Results in Using Attention for Increasing Attack Identification Efficiency",

abstract = "In previous work, we proposed an end-to-end intrusion early detection system to identify network attacks in real-time before they are complete and could cause more damage to the system under attack. To implement the approach, we have used a deep neural network which was trained in a supervised manner to extract relevant features from raw network traffic in order to classify network flows into different types of attacks. In this work, we discuss the initial results of the benefits that an attention mechanism brings to the classification performance and the capacity of the network to detect attacks earlier. We empirically evaluate our approach on the CICIDS2017 dataset. Preliminary results show that the attention mechanism improves both the balanced accuracy of the classifier as well as the early detection of attacks.",

keywords = "Attention Mechanism, Deep learning, Convolution Neural Network, Early Intrusion Detection System",

author = "Tanwir Ahmad and Dragos Truscan and J{\"u}ri Vain",

year = "2023",

month = may,

doi = "10.1109/ICSTW58534.2023.00038",

language = "English",

isbn = "979-8-3503-3336-7",

series = "IEEE International Conference on Software Testing, Verification and Validation Workshops",

publisher = "IEEE",

pages = "159--164",

booktitle = "Proceedings - 2023 IEEE 16th International Conference on Software Testing, Verification and Validation Workshops, ICSTW 2023",

address = "United States",

note = "IEEE International Conference on Software Testing, Verification and Validation, ICST ; Conference date: 16-04-2023 Through 20-04-2023",

url = "https://conf.researchr.org/home/icst-2023",

}

Ahmad, T , Truscan, D & Vain, J 2023, Preliminary Results in Using Attention for Increasing Attack Identification Efficiency. in Proceedings - 2023 IEEE 16th International Conference on Software Testing, Verification and Validation Workshops, ICSTW 2023. IEEE International Conference on Software Testing, Verification and Validation Workshops, IEEE, pp. 159-164, IEEE International Conference on Software Testing, Verification and Validation, Dublin, Ireland, 16/04/23. https://doi.org/10.1109/ICSTW58534.2023.00038

Preliminary Results in Using Attention for Increasing Attack Identification Efficiency. / Ahmad, Tanwir ; Truscan, Dragos; Vain, Jüri.
Proceedings - 2023 IEEE 16th International Conference on Software Testing, Verification and Validation Workshops, ICSTW 2023. IEEE, 2023. p. 159-164 (IEEE International Conference on Software Testing, Verification and Validation Workshops).

Research output: Chapter in Book/Conference proceeding › Published conference proceeding › Scientific › peer-review

TY - GEN

T1 - Preliminary Results in Using Attention for Increasing Attack Identification Efficiency

AU - Ahmad, Tanwir

AU - Truscan, Dragos

AU - Vain, Jüri

N1 - Conference code: 16

PY - 2023/5

Y1 - 2023/5

N2 - In previous work, we proposed an end-to-end intrusion early detection system to identify network attacks in real-time before they are complete and could cause more damage to the system under attack. To implement the approach, we have used a deep neural network which was trained in a supervised manner to extract relevant features from raw network traffic in order to classify network flows into different types of attacks. In this work, we discuss the initial results of the benefits that an attention mechanism brings to the classification performance and the capacity of the network to detect attacks earlier. We empirically evaluate our approach on the CICIDS2017 dataset. Preliminary results show that the attention mechanism improves both the balanced accuracy of the classifier as well as the early detection of attacks.

AB - In previous work, we proposed an end-to-end intrusion early detection system to identify network attacks in real-time before they are complete and could cause more damage to the system under attack. To implement the approach, we have used a deep neural network which was trained in a supervised manner to extract relevant features from raw network traffic in order to classify network flows into different types of attacks. In this work, we discuss the initial results of the benefits that an attention mechanism brings to the classification performance and the capacity of the network to detect attacks earlier. We empirically evaluate our approach on the CICIDS2017 dataset. Preliminary results show that the attention mechanism improves both the balanced accuracy of the classifier as well as the early detection of attacks.

KW - Attention Mechanism

KW - Deep learning

KW - Convolution Neural Network

KW - Early Intrusion Detection System

U2 - 10.1109/ICSTW58534.2023.00038

DO - 10.1109/ICSTW58534.2023.00038

M3 - Published conference proceeding

SN - 979-8-3503-3336-7

T3 - IEEE International Conference on Software Testing, Verification and Validation Workshops

SP - 159

EP - 164

BT - Proceedings - 2023 IEEE 16th International Conference on Software Testing, Verification and Validation Workshops, ICSTW 2023

PB - IEEE

T2 - IEEE International Conference on Software Testing, Verification and Validation

Y2 - 16 April 2023 through 20 April 2023

ER -

Ahmad T , Truscan D, Vain J. Preliminary Results in Using Attention for Increasing Attack Identification Efficiency. In Proceedings - 2023 IEEE 16th International Conference on Software Testing, Verification and Validation Workshops, ICSTW 2023. IEEE. 2023. p. 159-164. (IEEE International Conference on Software Testing, Verification and Validation Workshops). doi: 10.1109/ICSTW58534.2023.00038

Preliminary Results in Using Attention for Increasing Attack Identification Efficiency

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Fingerprint

Cite this