A systematic Analysis: Molecular Information in viral Disease using Deep Learning Auto Encoder

ABSTRACT

Hundreds of millions of people around the world suffer from viral infections every year. However, some of them have neither vaccine nor effective treatment during and after viral infection. Such as pneumonia, severe acute respiratory syndrome type 2 (SARS -2), HIV infection and Hepatitis-C virus. These viral diseases also directly and indirectly cause cardiovascular disease (CVD). Recently, the Deep Neural Network (DNN)-assisted molecular interaction (information) (MI) transceiver (transmitter Tx, and receiver Rx) design was brought to the fore to break the issues of traditional molecular information (MI) inside and outside human body. In this paper, we use DNN based approach to design and implement a new transceiver (Tx/Rx). We investigate DNN-assisted MI- Tx/Rx, multilayer perception DNN auto-encoder (MLP-AE), and convolutional neural network auto-encoder (CNN-AE), respectively. We apply an MLP-AE and CNN-AE to simultaneously accomplish the task of modulation, demodulation, and equalization as a point-to-point scheme. © 2021 IEEE.