Enhancing Resilience in Next-Generation Wireless Networks Through Deep Learning for Security Enhancement

Safeguarding strong protection and efficient intrusion recognition are important reflections in the growth and protection of Next-Generation Wireless Networks. These original networks, considered by their enhanced rapidity, ability, and connectivity, are also enhanced by enlarged vulnerabilities. To discover this, a complete protection framework is very vital. In the realm of Next-Generation Wireless Networks, safety actions go beyond traditional protocols, demanding advanced strategies. A main segment of this security architecture is intrusion detection, a positive method to identify and answer to unauthorized entries, malicious actions, and potential assaults. Effective safety models namely deep learning (DL) for intrusion detection have been proposed. Many researches are mainly focused on developing DL approaches for superior safety in the system. This work improves the finest DL-based intrusion detection in the Next-Generation Wireless Network (ODLID-NGWN) method. The projected ODLID-NGWN tactic enables for recognition and identification of the occurrence of the intrusions in the system. To complete this, the ODLID-NGWN method monitors a linear scaling normalization methodology as a pre-processing phase. Also, the Farmland Fertility Feature Selection (F3S) method is chiefly employed for the assortment of feature subsets. For intrusion or malicious activity detection, a hybrid DL (HDL) model comprising a long short-term memory (LSTM) gated recurrent unit (GRU) model can be used. To improve the performance of the HDL model, a modified cheetah optimization (MCO) algorithm can be applied to the hyperparameter tuning procedure. The performance analysis of the ODLID-NGWN methodology is tested by employing a benchmark dataset. The experimental values stated that the ODLID-NGWN technique gains maximum detection performance over other models.