MS-ADS: Multistage Spectrogram image-based Anomaly Detection System for IoT security

Zeeshan Ahmad, Adnan Shahid Khan, Kartinah Zen, Farhan Ahmad

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)

Abstract

The innovative computing idea of Internet-of-Things (IoT) architecture has gained tremendous popularity over the last decade, resulting in an exponential increase in the connected devices and the data processed in the IoT networks. Since IoT devices collect a massive amount of sensitive information exchanged over the traditional internet, security has become a prime concern due to the more frequent generation of network anomalies. A network-based anomaly detection system can provide the much-needed efficient security solution to the IoT network by detecting anomalies at the network entry points through constant traffic monitoring. Despite enormous efforts by researchers, these detection systems still suffer from lower detection accuracy in detecting anomalies and generate a high false alarm rate and false-negative rate in classifying network traffic. To this end, this paper proposes an efficient Multistage Spectrogram image-based network Anomaly Detection System (MS-ADS) using a deep convolution neural network that utilizes a short-time Fourier Transform to transform flow features into spectrogram images. The results demonstrate that the proposed method achieves high detection accuracy of 99.98% with a reduction in the false alarm rate to 0.006% in classifying network traffic. Also, the proposed scheme improves predicting the anomaly instances by 0.75% to 4.82%, comparing the benchmark methodologies to exhibit its efficiency for the IoT network. To minimize the computational and training cost for the model re-training phase, the proposed solution demonstrates that only 40500 network flows from the dataset suffice to achieve a detection accuracy of 99.5%.

Original languageEnglish
Article numbere4810
Number of pages21
JournalTransactions on Emerging Telecommunications Technologies
Volume34
Issue number8
Early online date19 Jun 2023
DOIs
Publication statusPublished - 1 Aug 2023
Externally publishedYes

Cite this