Realistic performance model for vehicle-to-infrastructure communications

This paper introduces extensions of a previously proposed range-dependent modified Gilbert model for generation of realistic error patterns. With the proposed extensions our model can be used to generate signal-to-noise ratio trends corresponding to the error patterns for infrastructure-to-vehicle communications. The model parameters are estimated based on realistic highway measurements using off-the-shelf IEEE 802.11p devices. Finally, a scalar quantization-based framework for reduction of the model complexity is introduced and evaluated.