Cellular-V2X communications with weighted-power-based mode selection

Abstract

Cellular-assisted vehicular-to-everything (C-V2X) communication is a promising technology to enhance the road safety and support various infortainment services in future vehicular networks. However, the integration of vehicular communication with the conventional cellular network faces many new design challenges, as vehicles typically use different spectrum band (5.9 GHz) from the cellular systems, and they usually follow different distributions and have more stringent reliability and latency requirements, as compared with the conventional cellular users. In this paper, we study C-V2X communication where the V2X resource pool is managed by the eNodeBs and reused in different cells, with the cells divided according to the Voronoi tessellation of the eNodeBs. We investigate three connection modes for a vehicle to receive safety/entertainment information, namely, from its associated eNodeB, from a neighboring vehicle on the same road, or from a neighboring vehicle on a different road. We propose a weighted-power-based mode selection scheme which offers great flexibility to balance the load of the cellular network and improve the coverage probability. By using stochastic geometry, we derive the analytical expressions for the mode selection and coverage probabilities, which are validated by extensive simulations. Furthermore, by optimizing the weighting factors, the proposed mode selection scheme leads to significantly better performance than the conventional distance-based and maximum-power-based mode selection schemes.