Modeling for geometry-based massive MIMO V2V channels and analysis of space-time correlation

Abstract

This paper proposed a three-dimensional dual-spheres confocal ellipsoid geometry-based model for massive MIMO vehicle-to-vehicle (V2V) communication. In this proposed model, the received signal is composed of non-line-of-sight propagation components generated via effective scatterers and line-of-sight propagation components. According to the geometric relationships in the model, the information of angle and distance can be derived, and then the space-time correlation function expression is obtained. To calculate parameters, modified method of equal areas is employed, and the influence on space-time correlation of scatters in different environments is analyzed. Channel characteristics in different scenarios such as highway scenario, urban road scenario and tunnel scenario are compared and studied. Simulation results show that the scatterers distribution and angle information of antenna arrays make a joint contribution on spatial cross-correlation, and moreover, the degree of anisotropy of scatterers and multipath components have distinct impacts on temporal auto-correlation. Compared with the results of 2D model, the proposed 3D model in this paper is more appropriate to characterize V2V communication in various scenarios.