This paper presents a three-dimensional (3D) massive multiple-input and multiple-output (MIMO) antenna array model, which includes the spherical array assumption and geometric properties for future fifth generation (5G) wireless communications. A parametric approximation algorithm is developed for estimating the spatial fading correlations (SFCs) and channel capacities of the 3D massive MIMO antenna array systems under different power angular spectrum (PAS). The relationship between correlation with the spacing of antenna arrays and angular parameters was classified. The results show that the simulation values of the approximate method fit the theoretical calculation very well, thereby validating the feasibility of the proposed 3D large-scale massive MIMO model.

Introduction

Massive multiple-input and multiple-output (MIMO) technologies have been suggested as a promising technology in modern wireless communication systems [1]. The fifth generation (5G) wireless communication networks providing with more stable and high-speed wireless access services are expected to further improve user experience, which put forward higher requirements related to the artificial intelligence for MIMO key technologies [2]–[۴]. It has been demonstrated that the scattering distribution and MIMO antenna structures and polarization employed have a critical impact on the MIMO channel performance. To establish more superior communication systems, one essential work is to investigate the channel characteristics for 5G massive MIMO channel models [5]. As the most fundamental channel parameters, the spatial fading correlation (SFC) and channel capacity play an important role in the planning, evaluation, and optimization of massive MIMO wireless communication systems [6], [7]. Until now, many researchers have carried out works and modeling analysis of antenna arrays for the wireless transmission environment. References [8]–[۱۱] proposed a series of MIMO planar antenna arrays, such as the uniform linear array (ULA), uniform circular array (UCA), uniform rectangular array (URA), and the L-shaped array configurations are proposed and investigated. The correlation-based traditional Kronecker models were demonstrated to help analyze the performance of compact MIMO systems [12].