A Stacked Autoencoder and Multilayer Perceptrons for mmWave Beamforming Prediction

The millimeter-wave frequencies planned for 6G systems present challenges channel modeling. At these frequencies, surface roughness affects wave propagation and causes severe attenuation of (mmWave) signals. In general, beamforming techniques compensate this problem. Analog has some major advantages over its counterpart, digital beamforming, because it uses low-cost phase shifters massive MIMO compared to that provides more accurate faster results in determining user However, suffers from high complexity expensive design, making unsuitable mmWave systems. proposed so far analog are often challenging practice. work, we have a deep learning model beams training helps predict the optimal beam vector. Our an available dataset 18 base stations, 1 million users, 60 GHz frequency. process first applies stacked autoencoder extract features datasets, then multilayer perceptron (MLP) train beams. Then, evaluated by computing mean squared error between expected predicted using test set. show efficiency benchmark method, which only MLP process.