Transformer-Based Anomaly Detection in IoT Networks Using Multimodal Sensor Data

Abstract

This study investigates transformer-based anomaly detection in IoT networks using multimodal sensor data. As IoT environments expand across smart homes, healthcare, industry, and critical infrastructure, the ability to detect abnormal behaviour from continuous, heterogeneous data streams has become increasingly important. Conventional anomaly detection techniques often struggle with nonlinear relationships, temporal dependencies, missing values, and noisy sensor inputs, especially when data are collected from multiple modalities. To address these limitations, the study proposes a transformer-based framework that learns long-range dependencies and cross-modal interactions more effectively than traditional approaches. The model is designed to identify anomalies by integrating information from diverse sensor sources and capturing subtle deviations that may not be visible in individual data streams. The framework is expected to improve detection accuracy, reduce false alarms, and enhance robustness under real-world IoT conditions. This research contributes to the growing literature on intelligent IoT security and time-series analytics by demonstrating the potential of attention-based deep learning for multimodal anomaly detection. The findings are intended to support practical applications in fault diagnosis, predictive maintenance, and cyber-physical system monitoring.

REFERENCES

[1] M. Yassine and T. Flaus, "Anomaly Detection for Industrial Sensors Using Transformers," in 2023 IEEE International Conference on Future Internet of Things and Cloud (FiCloud), 2023, pp. 167–174.

[2] F. Zeng, M. Chen, C. Qian, Y. Wang, Y. Zhou, and W. Tang, "Multivariate time series anomaly detection with adversarial transformer architecture in the Internet of Things," Future Gener. Comput. Syst., vol. 144, pp. 244–255, 2023.

[3] S. Tuli, G. Casale, and N. R. Jennings, "TranAD: Deep Transformer Networks for Anomaly Detection in Multivariate Time Series Data," Proc. VLDB Endow., vol. 15, no. 6, pp. 1201–1214, 2022.

[4] L. Barbieri, M. Brambilla, M. Stefanutti, C. Romano, N. De Carlo, and M. Roveri, "A Tiny Transformer-Based Anomaly Detection Framework for IoT Solutions," IEEE Open J. Signal Process., vol. 4, pp. 234–245, 2023.

[5] "Improving the accuracy of Anomaly Detection in Multimodal Sensors," ACM Digit. Libr., 2024.

[6] R. C. Nwachukwu, "Cross-modal feature learning for multi-sensor IoT anomaly detection: a comprehensive empirical analysis of cyber physical systems," Appl. Comput. Inform., 2026.

[7] "AI-based intelligent sensing detection of cybersecurity threats using multimodal sensor data in smart devices," Sci. Rep., vol. 16, Art. no. 11091, 2026.

[8] "Anomaly detection for IoT-based smart buildings," Malmö University Repository, 2025.

[9] "IoT Graph Learning-based Anomaly and Intrusion Detection through multi-modal data fusion," in Proc. DATA Conf., 2024.

[10] L. Aversano, M. L. Bernardi, M. Cimitile, R. Pecori, and L. Veltri, "Effective Anomaly Detection Using Deep Learning in IoT Systems," Wireless Commun. Mobile Comput., vol. 2021, Art. no. 9054336, 2021.

[11] "Evaluating large transformer models for anomaly detection of resource-constrained IoT devices for intrusion detection system," Sci. Rep., 2025.

[12] "A Framework for Anomaly Detection in IoT," IEEE Xplore.

[13] I. Ahmed and M. Asif, “The Role of HR in Managing Quiet Quitting and Employee Disengagement in Gen Z Employees of Telecom Sector,” Policy Journal of Social Science Review, vol. 4, no. 6, pp. 118–151, 2026, doi: 10.5281/zenodo.20581688.

[14] M. Asif, M. Abid, and A. Riaz, “Psychological drivers of investment decision making: A multi bias analysis of an emerging market’s retail investors,” Contemporary Journal of Social Science Review, vol. 4, no. 2, pp. 677–688, 2026, doi: 10.63878/cjssr.v4i2.2608.

[15] R. D. A. Khan, H. Ping, and M. Asif, “The impact of green human resource management on employee green performance through green commitment and transformational leadership,” Center for Management Science Research, vol. 4, no. 5, pp. 635–677, 2026, doi: 10.5281/zenodo.20510765.

[16] M. Asif, S. Karim, A. Latif, H. A. H. Asim, and A. Kareem, “Impact of behavioural biases on investment decisions: A study of individual investors in Pakistan,” Contemporary Journal of Social Science Review, vol. 4, no. 1, pp. 1538–1550, 2026, doi: 10.63878/cjssr.v4i1.2578.

[17] D. Mohiuddin, A. A. Zaveri, I. Ahmed, and M. Umar, “A systematic literature review of multi-channel analytics linked to POS and connected to food businesses in the UK,” in 2026 International Conference on AI Innovations and Industry (ICAIII), 2026, pp. 1–6. doi: 10.1109/ICAIII69475.2026.11521642.

[18] D. Mohiuddin, M. H. Tariq, and A. Tahir, “The Impact of Generative AI on Personalized Content Marketing in E-Commerce,” Inverge Journal of Social Sciences, vol. 4, no. 1, pp. 162–188, 2025. doi: 10.63544/ijss.v4i1.288.

[19] M. Asif and M. Bashir, “Augmentation or Anxiety? The Mediating Role of Employee Trust in The Relationship Between Generative AI Implementation, Job Crafting, and Productivity,” The Critical Review of Social Sciences Studies, vol. 4, no. 1, pp. 4550–4583, 2026, doi: 10.59075/mrqkn978.