A Novel Data Offloading with Deep Learning Enabled Cyberattack Detection Model for Edge Computing in 6G Networks
Elsaid Md. Abdelrahim · 2022
Edge computing is the most fundamental drive force for enabling Beyond 5G (B5G) and 6G networks. Because of the unprecedented improvement from traffic volume and computation demand of future networks, multiaccess edge computing (MEC) has assumed a promising solution for providing cloud computing (CC) abilities from the radio access network (RAN) nearby end users. Now, the overview of deep learning (DL) and hardware technology offers a technique in identifying the present traffic status, data offloading, and cyberattack from MEC. This study develops a novel data offloading with DL enabled cyberattack detection (DADL-CAD) model for edge computing in 6G networks. The proposed DADL-CAD technique primarily designs recurrent neural network (RNN) model for traffic flow forecasting in the edge computing enabled 6G networks. In addition, adaptive sampling cross entropy (ASCE) model is utilized for maximizing the network efficiency by proper decision making connected to the offloading process. Moreover, competitive swarm optimization (CSO) with stacked autoencoder (SAE) method was executed for the detection of cyberattacks from the network. The performance validation of the DADL-CAD technique is examined under various aspects, and the comparative study reported the supremacy of the DADL-CAD technique over the recent approaches.