Enhancing Security in Physical Layer Communications Based on Continuous Wavelet Transform with Convolutional Vision Transformer

Nowadays, the Enhancing Security in Physical Layer Communications underway on implementing the 6G radio because the attacks are increasing more on wireless edges such as jamming and false based stations and it placed a low-cost software radio. However, in Cyber-Physical System (CPS) the unauthorized access affects system process, such information lacks on the system and there is no adverse effect about such data. The quick environment changes and unpredicted events are such various factor that affect CPS, and it gives false alarm on the detection of cyber-physical attack. To overcome these drawbacks an ensemble method combining Continuous Wavelet Transform (CWT) with Convolutional Vision Transformer (CVT) to enhance physical layer communication security. The process involves collecting Real-World Photoplethysmography (RW-PPG) dataset and then the preprocessing is done with the Principal Component Analysis (PCA) and applying CWT with CVT signal transformations. Classification is then performed using Automatic Modulation Classification (AMC) and Specific Emitter Identification (SEI). The proposed CWT with CVT model obtained high results in the context of precision (PPV), achieving a score of 0.692, Sensitivity (TPR) of 0.841, Specificity (TNR) of 0.985, and Area Under the Curve (AUC) of 0.827 when compared to existing methods like Recurrent Neural Network (RNN) respectively.