OPTIMIZED DEEP AUTOENCODER WITH L1 REGULARIZATION AND DROPOUT FOR ANOMALY DETECTION IN 6G NETWORK SLICING
DOI:
https://doi.org/10.33480/techno.v20i2.6912Keywords:
6g network slicing , anomaly detection , autoencoder , dropout , regularizationAbstract
The increasing complexity of 6G network slicing introduces new challenges in identifying abnormal behavior within highly virtualized and dynamic network infrastructures. This study aims to address the anomaly detection problem in 6G slicing environments by comparing the performance of three models: a supervised random forest classifier, a basic unsupervised autoencoder, and an optimized deep autoencoder enhanced with L1 regularization and dropout techniques. The optimized autoencoder is trained to reconstruct normal data patterns, with anomaly detection performed using a threshold- based reconstruction error approach. Reconstruction errors are evaluated across different percentile thresholds to determine the optimal boundary for classifying abnormal behavior. All models are tested on a publicly available 6G Network Slicing Security dataset. Results show that the optimized autoencoder outperforms both the baseline autoencoder and the random forest in terms of anomaly sensitivity. Specifically, the optimized model achieves an F1- score of 0.1782, a recall of 0.2095, and an accuracy of 0.714. These results indicate that introducing regularization and dropout significantly improves the ability of autoencoders to generalize and isolate anomalies, even in highly imbalanced datasets. This approach provides a lightweight and effective solution for unsupervised anomaly detection in next- generation network environments.
References
Alamsyah, N., Budiman, B., Setiana, E., Jennifer, V. C., & others. (2025). THE ROLE OF L1 REGULARIZATION IN ENHANCING LOGISTIC REGRESSION FOR EGG PRODUCTION PREDICTION. JITK (Jurnal Ilmu Pengetahuan Dan Teknologi Komputer), 10(4), 821–832.
Alamsyah, N., Fauzan, M. N., Putrada, A. G., & Pane, S. F. (2022). Autoencoder image denoising to increase optical character recognition performance in text conversion. 2022 International Conference on Advanced Creative Networks and Intelligent Systems (ICACNIS), 1–6.
Alamsyah, N., Kurniati, A. P., & others. (2024a). Airfare Fluctuation Analysis with Event and Sentiment Features by Stacking Ensemble Model. 2024 Ninth International Conference on Informatics and Computing (ICIC), 1–6.
Alamsyah, N., Kurniati, A. P., & others. (2024b). Event Detection Optimization Through Stacking Ensemble and BERT Fine-tuning For Dynamic Pricing of Airline Tickets. IEEE Access.
Allaw, Z., Zein, O., & Ahmad, A.-M. (2025). Cross- Layer Security for 5G/6G Network Slices: An SDN, NFV, and AI-Based Hybrid Framework. Sensors, 25(11), 3335.
Altalhan, M., Algarni, A., & Alouane, M. T.-H. (2025). Imbalanced Data problem in Machine Learning: A review. IEEE Access.
Asad, M., Ullah, I., Hafeez, M. A., Sistu, G., & Madden,
M. G. (2025). A Probabilistic Adversarial Autoencoder for Novelty Detection: Leveraging Lightweight Design and Reconstruction Loss. IEEE Access.
Ayano, K. (2024). Deep Learning-Based Anomaly Detection in TLS Encrypted Traffic. Proceedings of the Future Technologies Conference, 249–270.
Ming, Z., Yu, H., & Taleb, T. (2024). User Request Provisioning Oriented Slice Anomaly Prediction and Resource Allocation in 6G Networks. ICC 2024-IEEE International Conference on Communications, 3640– 3645.
Mirzakhaninafchi, H. (2024). Comparative Analysis of Deep Learning-Based Anomaly Detection Models for Gps Spoofing Detection [Master’s Thesis]. South Dakota State University.
Putrada, A. G., Alamsyah, N., Fauzan, M. N., & Oktaviani, I. D. (2024). Pearson Correlation for Efficient Network Anomaly Detection with Quantization on the UNSW-NB15 Dataset. 2024 International Conference on ICT for Smart Society (ICISS), 1–6.
Putrada, A. G., Alamsyah, N., Oktaviani, I. D., & Fauzan, M. N. (2024). LSTM For Web Visit Forecasting with Genetic Algorithm and Predictive Bandwidth Allocation. 2024 International Conference on Information Technology Research and Innovation (ICITRI), 53–58.
Rodríguez-Ossorio, J. R., Morán, A., Fuertes, J. J., Prada, M. A., Díaz, I., & Domínguez, M. (2025). Adaptive model based on ESN for anomaly detection in industrial systems. Evolving Systems, 16(1), 25.
Rullo, A., Alam, F., & Serra, E. (2025). Trace Encoding Techniques for Multi-Perspective Process Mining: A Comparative Study. Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, 15(1), e1573.
Singh, P., Pranav, P., & Dutta, S. (2025). Bi-GAN-LDA for cybersecurity: A hybrid deep learning framework for advanced network anomaly detection. Engineering Research Express, 7(2), 025238.
Tera, S. P., Chinthaginjala, R., Pau, G., & Kim, T. H. (2024). Towards 6G: An Overview of the Next Generation of Intelligent Network Connectivity. IEEE Access.
Walczyna, T., Jankowski, D., & Piotrowski, Z. (2024). Enhancing Anomaly Detection Through Latent Space Manipulation in Autoencoders: A Comparative Analysis. Applied Sciences, 15(1), 286.
Wei, K., Zhao, R., Kou, H., Chen, P., Cao, Y., Zheng, Y., & Deng, L. (2025). Dimensionality reduction of rolling bearing fault data based on graph-embedded semi- supervised deep auto-encoders. Engineering Applications of Artificial Intelligence, 152, 110689.
Zeng, G.-Q., Yang, Y.-W., Lu, K.-D., Geng, G.-G., & Weng, J. (2025). Evolutionary Adversarial Autoencoder for Unsupervised Anomaly Detection of Industrial Internet of Things. IEEE Transactions on Reliability
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 Valencia Claudia Jennifer Kaunang, Nur Alamsyah, Titan Parama Yoga, Acep Hendra, Budiman Budiman
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
The copyright of any article in the TECHNO Nusa Mandiri Journal is fully held by the author under the Creative Commons CC BY-NC license. The copyright in each article belongs to the author. Authors retain all their rights to published works, not limited to the rights set out on this page. The author acknowledges that Techno Nusa Mandiri: Journal of Computing and Information Technology (TECHNO Nusa Mandiri) is the first to publish with a Creative Commons Attribution 4.0 International license (CC BY-NC). Authors can enter articles separately, manage non-exclusive distribution, from manuscripts that have been published in this journal into another version (for example: sent to author affiliation respository, publication into books, etc.), by acknowledging that the manuscript was published for the first time in Techno Nusa Mandiri: Journal of Computing and Information Technology (TECHNO Nusa Mandiri); The author guarantees that the original article, written by the stated author, has never been published before, does not contain any statements that violate the law, does not violate the rights of others, is subject to the copyright which is exclusively held by the author. If an article was prepared jointly by more than one author, each author submitting the manuscript warrants that he has been authorized by all co-authors to agree to copyright and license notices (agreements) on their behalf, and agrees to notify the co-authors of the terms of this policy. Techno Nusa Mandiri: Journal of Computing and Information Technology (TECHNO Nusa Mandiri) will not be held responsible for anything that may have occurred due to the author's internal disputes.
