DPFT: A blockchain-enabled differentially private federated transformer for trust evaluation in vehicular networks

Wajahat Ali; Fahd Raza; Arshad Iqbal; Abdul Wadood; Herie Park; Hani Albalawi; Byung O Kang; Wajahat Ali; Fahd Raza; Arshad Iqbal; Abdul Wadood; Herie Park; Hani Albalawi; Byung O Kang

doi:10.3934/math.20261146

AIMS Mathematics

2026, Volume 11, Issue 9: 28801-28828. doi: 10.3934/math.20261146

Previous Article Next Article

Research article Special Issues

DPFT: A blockchain-enabled differentially private federated transformer for trust evaluation in vehicular networks

1.
School of Computing Sciences, Pak-Austria Fachhochschule, Haripur 22620, Pakistan
2.
Zero Emission Technologies Innovation Center, University of Tabuk, Tabuk, Saudi Arabia
3.
Department of Electrical Engineering, Faculty of Engineering, University of Tabuk, Saudi Arabia
4.
Department of Electrical Engineering, Dong-A University, Busan, Republic of Korea
5.
Department of ICT Integrated Safe Ocean Smart Cities Engineering, Dong-A University, Busan, Republic of Korea

Received: 13 April 2026 Revised: 20 July 2026 Accepted: 24 July 2026 Published: 08 September 2026
MSC : 68P27, 68T07, 68W15

Intelligent transportation systems (ITS) rely on distributed data generated by connected vehicles, roadside infrastructure, and edge sensors. However, malicious participants may inject unreliable information that threatens transportation safety and operational stability. This study proposes a differentially private federated transformer (DPFT) framework for privacy-preserving node trust classification in vehicular networks. The framework combines Swin transformer-based feature learning, federated training, gradient clipping, Gaussian-noise-based differential privacy, and trimmed mean aggregation. A private blockchain records the resulting trust scores to provide tamper-resistant auditing and decentralized accountability. Experimental results show that DPFT achieves an area under the receiver operating characteristic curve (AUROC) of 0.90, an F1 score of 0.77, and more than 78% classification accuracy. It outperforms FedAvg-long short-term memory (LSTM) and FedAvg-Trans and maintains over 70% accuracy when 25% of the participating nodes are Byzantine. The blockchain component processes 55 transactions per second with a latency of 248 milliseconds, demonstrating practical scalability under the evaluated federated simulation setting.
- blockchain,
- differential privacy,
- federated learning,
- Swin transformer,
- vehicular networks,
- trust evaluation
Citation: Wajahat Ali, Fahd Raza, Arshad Iqbal, Abdul Wadood, Herie Park, Hani Albalawi, Byung O Kang. DPFT: A blockchain-enabled differentially private federated transformer for trust evaluation in vehicular networks[J]. AIMS Mathematics, 2026, 11(9): 28801-28828. doi: 10.3934/math.20261146

Related Papers:

Abstract

Intelligent transportation systems (ITS) rely on distributed data generated by connected vehicles, roadside infrastructure, and edge sensors. However, malicious participants may inject unreliable information that threatens transportation safety and operational stability. This study proposes a differentially private federated transformer (DPFT) framework for privacy-preserving node trust classification in vehicular networks. The framework combines Swin transformer-based feature learning, federated training, gradient clipping, Gaussian-noise-based differential privacy, and trimmed mean aggregation. A private blockchain records the resulting trust scores to provide tamper-resistant auditing and decentralized accountability. Experimental results show that DPFT achieves an area under the receiver operating characteristic curve (AUROC) of 0.90, an F1 score of 0.77, and more than 78% classification accuracy. It outperforms FedAvg-long short-term memory (LSTM) and FedAvg-Trans and maintains over 70% accuracy when 25% of the participating nodes are Byzantine. The blockchain component processes 55 transactions per second with a latency of 248 milliseconds, demonstrating practical scalability under the evaluated federated simulation setting.

References

[1]	N. Dasanayaka, Y. M. Feng, Analysis of vehicle location prediction errors for safety applications in cooperative-intelligent transportation systems, IEEE Trans. Intell. Transp. Syst., 23 (2022), 15512–15521. https://doi.org/10.1109/TITS.2022.3141710 doi: 10.1109/TITS.2022.3141710
[2]	S. Misra, C. Roy, R. Ghosh, P-VERSE: Prioritization of vehicles to enhance road safety in IoT environment, IEEE Int. Things J., 10 (2023), 16099–16106. https://doi.org/10.1109/JIOT.2023.3267182 doi: 10.1109/JIOT.2023.3267182
[3]	Y. C. Fu, C. L. Li, F. R. Yu, T. H. Luan, Y. Zhang, A survey of driving safety with sensing, vehicular communications, and artificial intelligence-based collision avoidance, IEEE Trans. Intell. Transp. Syst., 23 (2022), 6142–6163. https://doi.org/10.1109/TITS.2021.3083927 doi: 10.1109/TITS.2021.3083927
[4]	P. Rani, R. Sharma, Intelligent transportation system for internet of vehicles based vehicular networks for smart cities, Comput. Electr. Eng., 105 (2023), 108543. https://doi.org/10.1016/j.compeleceng.2022.108543 doi: 10.1016/j.compeleceng.2022.108543
[5]	L. J. Zhou, W. Tu, Q. Q. Li, D. J. Guan, A heterogeneous streaming vehicle data access model for diverse IoT sensor monitoring network management, IEEE Int. Things J., 11 (2024), 26929–26943. https://doi.org/10.1109/JIOT.2024.3384493 doi: 10.1109/JIOT.2024.3384493
[6]	S. A. Alhandi, H. Kamaludin, N. A. Mohammed, Trust evaluation model in IoT environment: a comprehensive survey, IEEE Access, 11 (2023), 11165–11182. https://doi.org/10.1109/ACCESS.2023.3240990 doi: 10.1109/ACCESS.2023.3240990
[7]	U. Ahmed, I. Raza, S. A. Hussain, Trust evaluation in cross-cloud federation: survey and requirement analysis, ACM Comput. Surv. (CSUR), 52 (2019), 1–37. https://doi.org/10.1145/3292499 doi: 10.1145/3292499
[8]	J. Zhao, J. F. Huang, N. X. Xiong, An effective exponential-based trust and reputation evaluation system in wireless sensor networks, IEEE Access, 7 (2019), 33859–33869. https://doi.org/10.1109/ACCESS.2019.2904544 doi: 10.1109/ACCESS.2019.2904544
[9]	M. Chen, Y. X. Zuo, X. Y. Jia, Y. Liu, X. H. Yu, K. Zheng, CEM: A convolutional embedding model for predicting next locations, IEEE Trans. Intell. Transp. Syst., 22 (2021), 3349–3358. https://doi.org/10.1109/TITS.2020.2983647 doi: 10.1109/TITS.2020.2983647
[10]	Z. B. Ying, S. L. Cao, X. M. Liu, Z. Ma, J. F. Ma, R. H. Deng, PrivacySignal: Privacy-preserving traffic signal control for intelligent transportation system, IEEE Trans. Intell. Transp. Syst., 23 (2022), 16290–16303. https://doi.org/10.1109/TITS.2022.3149600 doi: 10.1109/TITS.2022.3149600
[11]	M. Usman, M. A. Jan, A. Jolfaei, SPEED: a deep learning assisted privacy-preserved framework for intelligent transportation systems, IEEE Trans. Intell. Transp. Syst., 22 (2021), 4376–4384. https://doi.org/10.1109/TITS.2020.3031721 doi: 10.1109/TITS.2020.3031721
[12]	D. Helbing, D. Brockmann, T. Chadefaux, K. Donnay, U. Blanke, O. Woolley-Meza, et al., Saving human lives: What complexity science and information systems can contribute, J. Stat. Phys., 158 (2015), 735–781. https://doi.org/10.1007/s10955-014-1024-9 doi: 10.1007/s10955-014-1024-9
[13]	X. Chen, J. Ding, Z. Y. Lu, A decentralized trust management system for intelligent transportation environments, IEEE Trans. Intell. Transp. Syst., 23 (2022), 558–571. https://doi.org/10.1109/TITS.2020.3013279 doi: 10.1109/TITS.2020.3013279
[14]	L. J. Wei, Y. H. Yang, J. Wu, C. N. Long, B. Li, Trust management for Internet of Things: a comprehensive study, IEEE Int. Things J., 9 (2022), 7664–7679. https://doi.org/10.1109/JIOT.2021.3139989 doi: 10.1109/JIOT.2021.3139989
[15]	P. Rani, C. Sharma, J. V. N. Ramesh, S. Verma, R. Sharma, A. Alkhayyat, Federated learning-based misbehavior detection for the 5G-enabled Internet of Vehicles, IEEE Trans. Consum. Electr., 70 (2024), 4656–4664. https://doi.org/10.1109/TCE.2023.3328020 doi: 10.1109/TCE.2023.3328020
[16]	Q. F. Lai, C. Xiong, J. Chen, W. Wang, J. X. Chen, T. R. Gadekallu, Improved transformer-based privacy-preserving architecture for intrusion detection in secure V2X communications, IEEE Trans. Consum. Electr., 70 (2024), 1810–1820. https://doi.org/10.1109/TCE.2023.3324081 doi: 10.1109/TCE.2023.3324081
[17]	A. Mohammed, Building trust in driverless technology: overcoming cybersecurity challenges, 2025, submitted for publication. https://doi.org/10.22541/au.173991300.08356728/v1
[18]	M. Mylrea, N. Robinson, Artificial intelligence (AI) trust framework and maturity model: applying an entropy lens to improve security, privacy, and ethical AI, Entropy, 25 (2023), 1429. https://doi.org/10.3390/e25101429 doi: 10.3390/e25101429
[19]	B. Qolomany, I. Mohammed, A. Al-Fuqaha, M. Guizani, J. Qadir, Trust-based cloud machine learning model selection for industrial IoT and smart city services, IEEE Int. Things J., 8 (2021), 2943–2958. https://doi.org/10.1109/JIOT.2020.3022323 doi: 10.1109/JIOT.2020.3022323
[20]	T. R. Ramesh, M. Vijayaragavan, M. Poongodi, M. Hamdi, H. H. Wang, S. Bourouis, Peer-to-peer trust management in intelligent transportation system: an Aumann's agreement theorem based approach, ICT Express, 8 (2022), 340–346. https://doi.org/10.1016/j.icte.2022.02.004 doi: 10.1016/j.icte.2022.02.004
[21]	A. M. Yang, B. S. Xie, Y. K. Liu, L. Y. Wang, J. Li, A network security situation prediction for consumer data in the Internet of Things using variational mode decomposition (VMD) and fused CNN-BiLSTM-attention, IEEE Trans. Consum. Electr., 70 (2024), 1122–1133. https://doi.org/10.1109/TCE.2023.3323546 doi: 10.1109/TCE.2023.3323546
[22]	H. T. Liu, H. F. Wang, Real-time anomaly detection of network traffic based on CNN, Symmetry, 15 (2023), 1205. https://doi.org/10.3390/sym15061205 doi: 10.3390/sym15061205
[23]	B. I. Hairab, H. K. Aslan, M. S. Elsayed, A. D. Jurcut, M. A. Azer, Anomaly detection of zero-day attacks based on CNN and regularization techniques, Electronics, 12 (2023), 573. https://doi.org/10.3390/electronics12030573 doi: 10.3390/electronics12030573
[24]	Y. Y. Wei, J. Jang-Jaccard, W. Xu, F. Sabrina, S. Camtepe, M. Boulic, LSTM-autoencoder-based anomaly detection for indoor air quality time-series data, IEEE Sensors J., 23 (2023), 3787–3800. https://doi.org/10.1109/JSEN.2022.3230361 doi: 10.1109/JSEN.2022.3230361
[25]	R. Shrestha, M. Mohammadi, S. Sinaei, A. Salcines, D. Pampliega, R. Clemente, et al., Anomaly detection based on LSTM and autoencoders using federated learning in smart electric grid, J. Parallel Distrib. Comput., 193 (2024), 104951. https://doi.org/10.1016/j.jpdc.2024.104951 doi: 10.1016/j.jpdc.2024.104951
[26]	G. B. Giannakis, Q. Ling, G. Mateos, I. D. Schizas, H. Zhu, Decentralized learning for wireless communications and networking, In: Splitting methods in communication, imaging, science, and engineering, Cham: Springer, 2016,461–497. https://doi.org/10.1007/978-3-319-41589-5_14
[27]	W. P. B. Lim, N. C. Luong, D. T. Hoang, Y. T. Jiao, Y. C. Liang, Q. Yang, Federated learning in mobile edge networks: a comprehensive survey, IEEE Commun. Surv. Tutor., 22 (2020), 2031–2063. https://doi.org/10.1109/COMST.2020.2986024 doi: 10.1109/COMST.2020.2986024
[28]	A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, et al., Attention is all you need, In: NIPS'17: Proceedings of the 31st International Conference on Neural Information Processing Systems, Red Hook, NY, United States: Curran Associates Inc., 2017, 6000–6010.
[29]	X. C. Liu, W. Liu, H. D. Ma, H. Y. Fu, Large-scale vehicle re-identification in urban surveillance videos, In: 2016 IEEE International Conference on Multimedia and Expo (ICME), Seattle, WA, USA, 2016, 1–6. https://doi.org/10.1109/ICME.2016.7553002
[30]	M. Maqsood, S. Yasmin, S. Gillani, M. Bukhari, S. Rho, S. S. Yeo, An efficient deep learning-assisted person re-identification solution for intelligent video surveillance in smart cities, Front. Comput. Sci., 17 (2023), 174329. https://doi.org/10.1007/s11704-022-2050-4 doi: 10.1007/s11704-022-2050-4
[31]	B. L. Jiao, L. Yang, L. Y. Gao, P. Wang, S. Z. Zhang, Y. N. Zhang, Vehicle re-identification in aerial images and videos: dataset and approach, IEEE Trans. Circuits Syst. Video Technol., 34 (2024), 1586–1603. https://doi.org/10.1109/TCSVT.2023.3298788 doi: 10.1109/TCSVT.2023.3298788

Reader Comments

Your name:*

Email:*
© 2026 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)