Linear Behaviour Modelling of SBB5089Z Power Amplifier Module with Data Mining

Abstract

In this study, firstly s-parameters of SBB5089Z Power Amplifier (PA) module, which could be used for the WiMAX systems at 3.5GHz and 5.8GHz have been measured according to changing DC conditions [VDC=0V:0.075V:5V] at 3.4GHz-3.6GHz & 5.7GHz-5.9GHz frequencies. Data obtained from measurement results have been modelled separately with Artificial Neural Network (ANN), Sequential Minimal Optimization Regression (SMOReg), k-Nearest Neighbour (k-NN), K-Star (KS) algorithms. The successes of the models have been compared according to the Coefficient of determination (R2), Root Mean Squared Error (RMSE), and Mean Absolute Error (MAE). According to obtained results it has been shown that k-NN, ANN and K-star algorithms could be successfully used in modelling linear performance parameters S11, S12, S21, S22 of SB1150Z PA module.