A General 3D Non-Stationary GBSM for Underground Vehicular Channels

Reliable and efficient communications are indispensable for vehicles in underground environments. Underground wireless channels present a number of unique properties, such as guided propagation, rich scatterers, and near-field/far-field effect. In this paper, a three-dimensional (3D) twin cluster geometry-based stochastic model (GBSM) is proposed to describe underground vehicular channel characteristics in both pillar and tunnel scenarios. The proposed model supports arbitrary trajectory mobility of vehicles and multiple antenna configurations at the transmitter (Tx) and receiver (Rx). The cluster time evolution is modeled by different scatterer generation and updating methods to simulate the channel characteristics such as non-stationarity, near-field/far-field differences, and waveguide effects. Based on the proposed channel model, the statistical characteristics of the channel are derived and simulated, including temporal autocorrelation function (ACF), spatial cross-correlation function (CCF), delay power spectrum density (PSD), Doppler PSD, etc. Besides, underground channel measurements at 2.5/3.5 GHz are conducted in a garage scenario. Comparison results of channel measurements and simulations validate the accuracy and usefulness of the proposed GBSM.