Analytical Performance Evaluation of Full-Dimensional MIMO Systems Using Realistic Spatial Correlation Models

In this paper, we compute the analytical spatial channel correlation matrix in a polarized uniform planar array multiple-input multiple-output (MIMO) system. The antennas are considered with three-dimensional gain patterns. A realistic multiclustered propagation environment is considered. Earlier works in this area assumed arbitrary spatial clustering parameters, which do not adhere to practical situations. On the other hand, practical channel models provide methods for generating channel coefficients through simulation only. In this paper, cluster parameters from standard models such as the 3rd Generation Partnership Project (3GPP) WINNER II MIMO channel model for long term evolution (LTE)-Advanced and other proposed models for future SC systems for sub-6 GHz and mmWave frequencies have been used to arrive at an analytical correlation coefficient expression. The spatial correlation has been shown to depend on the parameters of the cluster power angular spectrum (PAS). The analytical performance metrics such as bit error rate (BER) and spectral efficiency (SE) arrived using the correlation matrices are found to match closely with those obtained from the Monte Carlo simulations using the 3GPP channel models. We also show the difference in SE results obtained using the proposed model and a model based on an existing hypothetical PAS assumption whereby the usefulness of the proposed model for designing future SC systems is reflected. Also, the analytical method developed here results in significant computation time reduction for calculation of different metrics such as BER and SE.