Performance of Massive MIMO Systems Considering Three Dimensional Array-based Spatially Correlated Channel

Abstract

This work deal with the uplink phase of the pilot-based channel estimation (CE) process for a Massive multiple-input multiple-output (M-MIMO) network operating under the time-division duplex protocol and considering spatially correlated channels. Additionally, this work investigates and examines uncorrelated channels and CC. In addition, the channels’s spatial correlation is investigated using two models, namely, the local scattering model (LSM) and the exponential correlation model (ECM), where each model is examined for two array configurations (i.e., uniform linear array (ULA) and uniform planar array (UPA)). Furthermore, we propose a UPA configuration for each model based on the Kronecker product (KP) of the ULA configuration, i.e., we propose the ECM and LSM for the UPA configuration, respectively, based on the KP and using either the ECM or the LSM for the ULA configuration. A comparison and evaluation of these models and their influence on the CE process for a M-MIMO network is analyzed. Numerical results are offred to assert our theoretical expressions.