Deep Learning Recurrent Neural Network based Wireless Power Allocation for Hybrid TDMA-NOMA System

Recurrent Neural Networks (RNNs) come under the category of Artificial Neural Networks (ANNs), which usually make use of sequential data. This research study considers the problem of non-orthogonal multiple access (NOMA) systems. The main disadvantage of the existing system is that it causes a poor channel state error, although it has the advantage of perfect transmission of the signals by eliminating the problem of interference with the help of successive interference cancellation. Considering that the channel state information (CSI) is imperfect, as already mentioned above, this study uses the deep learning method to get perfect channel state information. Thus, the analysis shows that the performance of the NOMA system under imperfect CSI is 0. This project is carried out using MATLAB R2021r software.