연구 분야: Infrastructure
학회: Cluster Computing
6G wireless will be the new breakthrough in telecommunications technology, providing speeds that are unprecedented, ultra-low latency, and massive connectivity. However, the growth of these networks has given way to growing security challenges against them. The enhanced complexity and scale of 6G systems introduce new vulnerabilities and attack vectors; hence, efficient security solutions with adaptiveness are much needed to protect against such sophisticated cyber threats. This paper proposes a new security model designed for 6G networks based on these key security concerns: resilient probabilistic adversarial network (Resi-PAN). It embeds new methodologies to boost the intrusion detection and classification capabilities. We have used the hybrid differential elephant clan optimization algorithm in this paper for the optimal computation of regularization parameter λ, which improves the performance and generalization capability of the model. The Resi-PAN model itself fuses probabilistic adversarial networks with resilient optimization techniques to set itself apart with better adaptiveness and accuracy in the detection of most forms of intrusions. The invention is in the putting together of these forefront techniques into a robust adaptive security framework. The Resi-PAN model puts recurrent neural networks and adversarial training into service in an effort to render efficient processing of the dynamic and fast-changing network threats. Experiments performed on UNSW-NB 15, NSL-KDD, and CICIDS 2019 datasets show that the Resi-PAN model has outperformed other techniques existing in terms of accuracy, precision, recall, and F1-score, along with computational efficiency. This paper is therefore timely in presenting a major milestone in network security by offering a practical and effective solution for the emerging challenge of securing 6G wireless networks. In this regard, the proposed model serves current security needs and opens up future research and development to guard the next generation of communications.
| 발행 연도 | 2025년 |
|---|---|
| 인용수 | 0 |
| 출판 국가 | Andorra |
| 사이트 | Springer |
| 좋아요 수 | 0 |