Leveraging self attention driven gated recurrent unit with crocodile optimization algorithm for cyberattack detection using federated learning framework.

Journal: Scientific reports
Published Date: Jul 3, 2025

Abstract

Cybersecurity has been defined as a vital part of the developments, which is mainly related to technology. Enlarged cybersecurity safeguards that the data remains safe. Cyberattacks like computer malware, denial-of-service (DoS) attacks, or unauthorized access led to severe damage and economic losses in large-scale systems. Cybersecurity includes decreasing the risk of mischievous computer, software, and network attacks. Novel techniques have been combined into emerging artificial intelligence (AI) that attains cybersecurity. ML is usually reflected as a sub-branch of AI, which is closely linked to data mining, computational statistics, and data science (DS) and mainly concentrates on generating computers to acquire data. Federated learning (FL) is one of the ML models that permits tackling cyberattack issues like security, data privacy, and access rights. This study proposes a Self-Attention Mechanism-Driven Federated Learning for Secure Cyberattack Detection with Crocodile Optimization Algorithm (SAMFL-SCDCOA) methodology. The main objective of the SAMFL-SCDCOA methodology is to provide an effective method for preventing cyberattacks in real time using FL and advanced optimization algorithms. Initially, the Z-score normalization is utilized to scale and standardize data to improve analysis consistency and accuracy. Furthermore, the feature selection (FS) process uses the crocodile optimization algorithm (COA) model. The proposed SAMFL-SCDCOA approach employs the gated recurrent unit with a self-attention (GRU-SA) model for the cybersecurity classification. Finally, the improved pelican optimization algorithm (IPOA) optimally adjusts the hyperparameter values of the GRU-SA model, leading to enhanced classification performance. A wide range of experiments has been accomplished to validate the performance of the SAMFL-SCDCOA technique under the CICIDS-2017 dataset. The comparison study of the SAMFL-SCDCOA technique emphasized a superior output of 99.04% over existing models.

Authors

  • Manal Abdullah Alohali
    Department of Information Systems, College of Computer and Information Sciences, Princess Nourah bint Abdulrahman University, P.O. Box 84428, Riyadh 11671, Saudi Arabia.
  • Hatim Dafaalla
    Department of Computer Science, Applied College at Mahayil, King Khalid University, Abha, Saudi Arabia.
  • Mohammed Baihan
    Department of Computer Science, Community College, King Saud University, P.O. Box 11451, Riyadh, Saudi Arabia.
  • Sultan Alahmari
    King Abdul Aziz City for Science and Technology (KACST), Cybersecurity Institute, Riyadh, Kingdom of Saudi Arabia.
  • Achraf Ben Miled
    Department of Computer Science, College of Science, Northern Border University, 73213, Arar, Saudi Arabia. ashraf.benmilad@nbu.edu.sa.
  • Othman Alrusaini
    Department of Engineering and Applied Sciences, Applied College, Umm Al-Qura University, Makkah, Saudi Arabia.
  • Ali Alqazzaz
    Faculty of Computing and Information Technology, University of Bisha, Bisha, Saudi Arabia.
  • Hanadi Alkhudhayr
    Department of Information Systems, Faculty of Computing and Information Technology, King Abdulaziz University, 25732, Rabigh, Saudi Arabia.

Keywords

No keywords available for this article.

External Resources

Popular Topics
Recent Journals