Optimized Hybrid Quantum-Classical Deep Learning Model for SYN Attack Detection in IoT Networks

Akshat Gaurav; Shin Hung Pan; Varsha Arya; Kowk Tai Chui; Brij B. Gupta

doi:10.1109/GCWkshps68340.2025.11591069

Optimized Hybrid Quantum-Classical Deep Learning Model for SYN Attack Detection in IoT Networks

Akshat Gaurav
, Shin Hung Pan
, Varsha Arya
, Kowk Tai Chui
, Brij B. Gupta

Electronic Engineering and Computer Science

Research output: Contribution to journal › Conference article › peer-review

Abstract

IoT networks are growing fast. But with that comes more risk. One serious threat is the SYN flood attack. It can shut down communication systems. To solve this, the paper proposed a hybrid model of deep learning and quantum computing. It integrates quantum computing and classical deep learning layers to detect SYN attacks. The model is tested on the CIC-DDoS2019 dataset. To make the model lightweight, SHIO is used for feature selection. In tests, the model reached 99.36% accuracy. Only 746 trainable parameters were used. That is much better than the state-of-the-art models.

Original language	English
Pages (from-to)	460-465
Number of pages	6
Journal	IEEE Globecom Workshops, GC Wkshps
Issue number	2025
DOIs	https://doi.org/10.1109/GCWkshps68340.2025.11591069
Publication status	Published - 2025
Event	2025 IEEE Globecom Workshops, GC Wkshps 2025 - Taipei, Taiwan, Province of China Duration: 8 Dec 2025 → 12 Dec 2025

Keywords

Hybrid Quantum-Classical Model
IoT Security
Quantum Machine Learning
SYN Attack Detection

Access to Document

10.1109/GCWkshps68340.2025.11591069

Cite this

@article{0f769fa489b94bfaa5ed23705b533e2e,

title = "Optimized Hybrid Quantum-Classical Deep Learning Model for SYN Attack Detection in IoT Networks",

abstract = "IoT networks are growing fast. But with that comes more risk. One serious threat is the SYN flood attack. It can shut down communication systems. To solve this, the paper proposed a hybrid model of deep learning and quantum computing. It integrates quantum computing and classical deep learning layers to detect SYN attacks. The model is tested on the CIC-DDoS2019 dataset. To make the model lightweight, SHIO is used for feature selection. In tests, the model reached 99.36\% accuracy. Only 746 trainable parameters were used. That is much better than the state-of-the-art models.",

keywords = "Hybrid Quantum-Classical Model, IoT Security, Quantum Machine Learning, SYN Attack Detection",

author = "Akshat Gaurav and Pan, \{Shin Hung\} and Varsha Arya and Chui, \{Kowk Tai\} and Gupta, \{Brij B.\}",

note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 2025 IEEE Globecom Workshops, GC Wkshps 2025 ; Conference date: 08-12-2025 Through 12-12-2025",

year = "2025",

doi = "10.1109/GCWkshps68340.2025.11591069",

language = "English",

pages = "460--465",

number = "2025",

}

TY - JOUR

T1 - Optimized Hybrid Quantum-Classical Deep Learning Model for SYN Attack Detection in IoT Networks

AU - Gaurav, Akshat

AU - Pan, Shin Hung

AU - Arya, Varsha

AU - Chui, Kowk Tai

AU - Gupta, Brij B.

PY - 2025

Y1 - 2025

N2 - IoT networks are growing fast. But with that comes more risk. One serious threat is the SYN flood attack. It can shut down communication systems. To solve this, the paper proposed a hybrid model of deep learning and quantum computing. It integrates quantum computing and classical deep learning layers to detect SYN attacks. The model is tested on the CIC-DDoS2019 dataset. To make the model lightweight, SHIO is used for feature selection. In tests, the model reached 99.36% accuracy. Only 746 trainable parameters were used. That is much better than the state-of-the-art models.

AB - IoT networks are growing fast. But with that comes more risk. One serious threat is the SYN flood attack. It can shut down communication systems. To solve this, the paper proposed a hybrid model of deep learning and quantum computing. It integrates quantum computing and classical deep learning layers to detect SYN attacks. The model is tested on the CIC-DDoS2019 dataset. To make the model lightweight, SHIO is used for feature selection. In tests, the model reached 99.36% accuracy. Only 746 trainable parameters were used. That is much better than the state-of-the-art models.

KW - Hybrid Quantum-Classical Model

KW - IoT Security

KW - Quantum Machine Learning

KW - SYN Attack Detection

UR - https://www.scopus.com/pages/publications/105045447109

U2 - 10.1109/GCWkshps68340.2025.11591069

DO - 10.1109/GCWkshps68340.2025.11591069

M3 - Conference article

AN - SCOPUS:105045447109

SN - 2166-0069

SP - 460

EP - 465

JO - IEEE Globecom Workshops, GC Wkshps

JF - IEEE Globecom Workshops, GC Wkshps

IS - 2025

T2 - 2025 IEEE Globecom Workshops, GC Wkshps 2025

Y2 - 8 December 2025 through 12 December 2025

ER -

Optimized Hybrid Quantum-Classical Deep Learning Model for SYN Attack Detection in IoT Networks

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this