A geometry based stochastic model for MIMO V2V channel simulation in cross-junction scenario

In this paper a multiple-input multiple-output (MIMO) simulation model for vehicle-to-vehicle (V2V) communication channels in an urban cross-junction scenario is presented. The model is an extension and modification of an existing T-junction model in the literature. Four propagation processes are considered: line-of-sight (LOS), single bounce reflections from side walls, double bounce reflections from side walls, and single bounce reflections from the corner of the building in front of the transmitter (TX) and receiver (RX). Each propagation process is linked to a cluster of scatterers, with a cluster size that varies with respect to the position of the TX and RX. The relations between angle-of-arrivals and angle-of-departures of all Multipath Components (MPCs) are derived depending on the different positions of the TX and RX. For the single bounce reflections from the corner a new method is being used where the scatterers are distributed randomly in a triangular plane, based on the assumption that corners of buildings typically have different scattering objects contributing to the reflection process. A complete expression for the time-variant transfer function is then derived by super positioning all contributions, including the LOS when this is available. The final results show that our model follows a realistic measurement based path-loss model, which subsequently makes the model a suitable candidate for analyzing MIMO V2V fading channels in cross-junction scattering environments.