Publication Date: 2024/02/07
Abstract: To demonstrate over the air transmission, it is essential to frame, alternate and execute a transmission systems repressed by neural networks. Autoencoders are used to train the entire system composed of transmitters and receivers. Estab- lishing a vital novel style of thinking regarding communications network design as a point to point regeneration task that seeks to optimise Tx and Rx systems into a single process by interpreting a transmission system as an autoencoder is performed. In this, several autoencoders such as deep encoder, convolutional autoencoder and a simplest possible autoencoder is simulated in Python. Lastly, BLER versus Eb/N0 for the (2,2) and (7,4) autoencoder is plotted.
Keywords: Autoencoder, Deep Learning, End-to-End Communication.
DOI: https://doi.org/10.5281/zenodo.10629098
PDF: https://ijirst.demo4.arinfotech.co/assets/upload/files/IJISRT24JAN1828.pdf
REFERENCES