Resource allocation in a MIMO network with smart reflective surfaces by using deep learning algorithms

Recently, wireless communications systems have developed significantly, and have impacted the spread of intelligent reflective surfaces (IRS) in multiple-input-multi-output (MIMO) networks. However, efficient allocation of resources such as power, bandwidth, and number of antennas among users in these networks remains a major challenge. A deep learning-relied (DL) algorithm for resource allocation in MIMO networks with IRS is suggested in this study. The proposed algorithm exploits the power of DL to optimize resource allocation in real-time, taking into account channel conditions as well as client motion. The performance of the suggested algorithm will be evaluated by performing multiple large-scale simulations and comparing it to existing resource allocation algorithms in MIMO/IRS system. The results show that the proposed LSTM-RNN/DL algorithm outperforms similar algorithms in terms of spectral efficiency, accuracy, and convergence speed. This research contributes to the development of resource allocation techniques in MIMO networks with IRS for outstanding efficiency using MATLAB program. Simulation results show BER of 104 with 99.8% performance efficiency to cover the increasing demand for wireless communication services.