Article: Transformer-based Modbus fuzz testing: a dual-strategy framework with self-adaptive flight firefly optimisation sampling and independent test case selection Journal: International Journal of Ad Hoc and Ubiquitous Computing (IJAHUC) 2025 Vol.50 No.1 pp.46 - 60 Abstract: In the era of increasing cyber threats, ensuring the security of industrial communication protocols such as Modbus is paramount. This paper proposes a novel fuzz testing framework for the Modbus TCP protocol that integrates a transformer-based test case generator, an improved firefly optimisation algorithm, and an effective test case selection mechanism. The transformer network learns complex sequential patterns to generate candidate test cases that accurately simulate real-world protocol interactions. Meanwhile, the enhanced firefly algorithm dynamically adjusts sampling probabilities - balancing exploration and exploitation - to reduce redundancy and increase the diversity of generated test cases. An independent selection mechanism further refines the candidates, filtering out duplicates and ensuring a high-quality, non-redundant test suite. Experimental evaluations demonstrate that our integrated approach significantly improves test coverage and vulnerability detection compared to traditional fuzz testing methods, offering a scalable solution to enhance the security of industrial control systems. Inderscience Publishers - linking academia, business and industry through research

Title: Transformer-based Modbus fuzz testing: a dual-strategy framework with self-adaptive flight firefly optimisation sampling and independent test case selection

Authors: Junlei Qian; Qingxin Dong; Kai Zeng; Xueqiang Du

Addresses: College of Electrical Engineering, North China University of Science and Technology, Hebei, 063210, China ' College of Electrical Engineering, North China University of Science and Technology, Hebei, 063210, China ' College of Electrical Engineering, North China University of Science and Technology, Hebei, 063210, China ' Tangshan Iron and Steel Enterprise Process Control and Optimization Technology Innovation Center, Tangshan ANODE Automation Company Ltd., Tangshan, Hebei, 063000, China

Abstract: In the era of increasing cyber threats, ensuring the security of industrial communication protocols such as Modbus is paramount. This paper proposes a novel fuzz testing framework for the Modbus TCP protocol that integrates a transformer-based test case generator, an improved firefly optimisation algorithm, and an effective test case selection mechanism. The transformer network learns complex sequential patterns to generate candidate test cases that accurately simulate real-world protocol interactions. Meanwhile, the enhanced firefly algorithm dynamically adjusts sampling probabilities - balancing exploration and exploitation - to reduce redundancy and increase the diversity of generated test cases. An independent selection mechanism further refines the candidates, filtering out duplicates and ensuring a high-quality, non-redundant test suite. Experimental evaluations demonstrate that our integrated approach significantly improves test coverage and vulnerability detection compared to traditional fuzz testing methods, offering a scalable solution to enhance the security of industrial control systems.

Keywords: industrial communication protocols; transformer; firefly algorithm; Modbus TCP; sequence optimisation.

DOI: 10.1504/IJAHUC.2025.148436

International Journal of Ad Hoc and Ubiquitous Computing, 2025 Vol.50 No.1, pp.46 - 60

Received: 03 Mar 2025
Accepted: 26 Apr 2025
Published online: 04 Sep 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: Transformer-based Modbus fuzz testing: a dual-strategy framework with self-adaptive flight firefly optimisation sampling and independent test case selection

Keep up-to-date