Deep Learning based MIMO detection in 6G Wireless Communication System

This paper proposes a deep learning algorithm to decode spheres in order to solve the multi-input multi-output (MIMO) receiver’s detection problem. The modified K and K1 sphere decoders achieve high performance in terms of spectral efficiency. This in turn will improve the data rate of next generation wireless communication systems. In order to reduce the computational complexity, the number of codewords needs to be reduced, which can be achieved with a deep learning algorithm based on a neural network. Simulation results depict the effectiveness of the proposed schemes in terms of BER, spectral efficiency, and computational time. Simulations were carried out using the MATLAB tool to support the proposed algorithm and validate the analytical results.