Geometry-Based Stochastic Channel Model for Vehicle-to-Vulnerable Road Users Communications in Critical Scenarios

Abstract

Direct vehicle-to-vulnerable road user (V2VRU) communication can prevent accidents by enabling mutual awareness between vehicles and vulnerable road users (VRUs) such as pedestrians and cyclists. A prerequisite for developing reliable V2VRU communication systems is having a channel model that captures the main propagation characteristics of this specific link type. In this work, a measurement-based parameterization and validation of a WINNER-type V2VRU geometry-based stochastic channel model (GSCM) is presented at 5.2 GHz, based on single-input single-output (SISO) wideband channel measurements in urban environments. The three most critical accident-prone scenarios involving pedestrians and cyclists are considered. The large-scale parameters (LSPs), path loss, shadow fading, delay spread, K-factor, and angular spreads, are estimated from the measurement data and fitted to log-normal distributions. The results reveal propagation characteristics that differ substantially from those reported for vehicle-to-vehicle (V2V) channels, including significantly higher path loss exponents under obstructed conditions and distinct LSP correlation structures. The proposed model is validated statistically and further assessed through link-level simulations. The simulations show that commonly used V2V channel models fail to reproduce the error floor observed in the measured V2VRU channels, which highlights the importance of using dedicated channel models when evaluating V2VRU safety applications.