TY - JOUR
T1 - LMT-SDNN
T2 - A Lightweight Malicious Traffic Detection Method for the Internet of Things Based on Multiteacher Distillation
AU - Xu, Dawei
AU - Liang, Yunfang
AU - Li, Chunhai
AU - Zhou, Xiaolong
AU - Zhang, Chuan
AU - Zhu, Liehuang
AU - Zhao, Jian
N1 - Publisher Copyright:
© 2014 IEEE. All rights reserved.
PY - 2026
Y1 - 2026
N2 - The rapid proliferation of the Internet of Things (IoT) devices, coupled with their inherent security vulnerabilities, has significantly expanded the attack surface, intensifying threats such as man-in-the-middle attacks, traffic hijacking, and distributed denial-of-service (DDoS) attacks, thereby posing serious risks to the security and reliability of the entire ecosystem. The network traffic associated with IoT devices is diverse and dynamic, often exhibiting complex structural features such as periodic fluctuations, varying packet sizes, and time-varying patterns that make detection challenging. Although deep learning has demonstrated strong capabilities in efficiently identifying complex and dynamic malicious traffic through powerful feature extraction and adaptive learning abilities, its high model complexity, substantial computational demands, and large parameter sizes hinder direct deployment on resource-constrained IoT devices. In order to tackle this issue, this article proposes a malicious traffic detection framework for the IoT based on multiteacher knowledge distillation (KD). The proposed model, termed the lightweight multiteacher spatiotemporal distillation neural network (LMT-SDNN), employs two high-performance teacher models: residual inception and a 1-D convolution neural network (1D-CNN) integrated with bidirectional long short-term memory (BiLSTM), to effectively capture the complex structural features of network traffic. Furthermore, a novel time-related window (TRW) loss function is design to enhance the student’s ability to capture temporal features, thereby improving its overall performance. The effectiveness of LMT-SDNN is validated through comparisons with five baseline models on two publicly available datasets, ToN _IoT and BoT_ IoT. Experimental results show that LMT-SDNN achieves a compression rate of over 99% in both model complexity and parameter count, while maintaining an accuracy exceeding 99%, indicating its strong potential for multiclass malicious traffic detection in IoT environments.
AB - The rapid proliferation of the Internet of Things (IoT) devices, coupled with their inherent security vulnerabilities, has significantly expanded the attack surface, intensifying threats such as man-in-the-middle attacks, traffic hijacking, and distributed denial-of-service (DDoS) attacks, thereby posing serious risks to the security and reliability of the entire ecosystem. The network traffic associated with IoT devices is diverse and dynamic, often exhibiting complex structural features such as periodic fluctuations, varying packet sizes, and time-varying patterns that make detection challenging. Although deep learning has demonstrated strong capabilities in efficiently identifying complex and dynamic malicious traffic through powerful feature extraction and adaptive learning abilities, its high model complexity, substantial computational demands, and large parameter sizes hinder direct deployment on resource-constrained IoT devices. In order to tackle this issue, this article proposes a malicious traffic detection framework for the IoT based on multiteacher knowledge distillation (KD). The proposed model, termed the lightweight multiteacher spatiotemporal distillation neural network (LMT-SDNN), employs two high-performance teacher models: residual inception and a 1-D convolution neural network (1D-CNN) integrated with bidirectional long short-term memory (BiLSTM), to effectively capture the complex structural features of network traffic. Furthermore, a novel time-related window (TRW) loss function is design to enhance the student’s ability to capture temporal features, thereby improving its overall performance. The effectiveness of LMT-SDNN is validated through comparisons with five baseline models on two publicly available datasets, ToN _IoT and BoT_ IoT. Experimental results show that LMT-SDNN achieves a compression rate of over 99% in both model complexity and parameter count, while maintaining an accuracy exceeding 99%, indicating its strong potential for multiclass malicious traffic detection in IoT environments.
KW - Internet of Things (IoT)
KW - knowledge distillation KD)
KW - lightweight neural network
KW - malicious traffic detection
UR - https://www.scopus.com/pages/publications/105029061810
U2 - 10.1109/JIOT.2026.3657416
DO - 10.1109/JIOT.2026.3657416
M3 - Article
AN - SCOPUS:105029061810
SN - 2327-4662
VL - 13
SP - 15665
EP - 15677
JO - IEEE Internet of Things Journal
JF - IEEE Internet of Things Journal
IS - 8
ER -