A Novel Geometric Cluster-Based Propagation-Graph Channel Model for Vehicle-to-Infrastructure Communications

A systematic vehicular network is the cornerstone of the future intelligent transportation system. In this paper, we propose a novel propagation-graph channel model based on geometric clusters. First, we propose the cluster-based propagation-graph channel model, which uses some geometric clusters to represent the surrounding environment. Then, a novel hard-core Poisson cluster process (PCP) is used to initialize and generate geometric clusters located within the maximal radio coverage. Regarding the cluster evolution, the cluster generation and recombination processes are influenced by transceivers’ velocities and distance-related link probability. Finally, the time-varying non-stationary vehicle-to-infrastructure (V2I) channel impulse response (CIR) is obtained. Based on the proposal, we emulate the V2I scenarios and investigate the cluster evolution, power delay profile (PDP), and Doppler power spectral density (DPSD), respectively. Relevant results reveal that our proposal can capture the channel non-stationarity of V2I channels, providing insights into the development of vehicular communications.