Improving Security for 6G Networks Using Beamforming Technology

Resumen

6G offers possibilities through technologies such as millimeter waves, terahertz waves, beamforming, and reconfigurable intelligent surfaces (RIS). Each technology introduces its own challenges. One of the most significant challenges is security, and this is the subject of our study. Vulnerabilities arise from narrow beams when using millimeter waves. Through this study, the security of 6G networks is enhanced with deep learning algorithms and using RIS to guide the smart beam to counter attacks. A multi-layered deep neural network (DNN) is used to predict the signal-to-noise ratio (SINR) and interference with and without the mobility of users. Additionally, a mathematical model was developed to optimize the SINR and counter some security threats, such as jamming, spoofing, and data poisoning. This work contains realistic channel modeling to simulate user mobility and logarithmic shading effects. Our results show that the maximum average SINR was 33.2351 using beamforming with RIS, and after adding deep learning, the improvement was 27.5%. Using a DNN model, SINR levels (25-45 dB) under three attacks and continuous user movement at a speed of , and a prediction accuracy of 0.032 was obtained for the root mean square error (RMSE) value. It should be noted that the achieved performance depends on the deployment and channel conditions considered. The proposed system provides a timely security response with a training period of 37 seconds, which is suitable for 6G requirements.