Pilot Overhead Reduction in 5G mmWave m-MIMO Systems Using Unsupervised and Supervised AI

In beam-forming systems, pilot signals play a critical role in achieving optimal channel estimation and mitigating interference, thereby ensuring data fidelity. However, this necessitates the transmission of pilot symbols, which occupy spectral resources and consequently reduce spectral efficiency (SE). To address this challenge, this work utilizes data-aided artificial intelligence (AI) tools for designing hybrid beam-forming systems maintaining SE with improving bit error rate (BER). The study investigates the application of supervised and unsupervised learning techniques to reduce the pilot overhead in 5G millimeter-wave (mm-Wave) massive MIMO hybrid beam-forming systems. The objective is to achieve pilot overhead reduction while maintaining same spectral efficiency and minimizing the bit error rate (BER). Results show that by applying the appropriate techniques, pilot overhead size is reduced by approximately 93% while maintaining and if possible improving system BER and SNR.