Antenna Design for MIMO SDR Systems Using Deep Learning

In this research, we present and put into practice a unique architecture for software defined radio (SDR) systems with multiple inputs multiple output (MIMO). The three following stages are used to develop the system: In order to achieve uplink interaction from customers to the ground station using time division duplex (TDD), a MIMO SDR communication platform must first be built. Using the DNN from our work, we then create a DNN decision server to help the MIMO SDR platform make smart antenna selection decisions, changing the performance tuning decision-making technique into an information driven decision-making technique. A norm-based antenna selection (NBAS) method is chosen for comparing in order to estimate the effectiveness of the DLBAS-MIMO SDR technology. The findings indicate that the applied DLBAS strategy outscored the MIMO system without AS with up to a 53% increase in average transmission power gain, and operated on par with the NBAS method in real-time.