A Qualitative Analysis of Vehicle Positioning Requirements for Connected Vehicle Applications

Many Connected Vehicle (CV) applications, including safety-critical ones such as collision warning, require lane-level positioning accuracy to function correctly. However, differential GNSS, the primary positioning method used by CVs in current deployments across the U.S., cannot always provide this level of accuracy. This is particularly true in urban environments. Alternative positioning methods or strategies must be developed to fill this gap. To determine what strategies are appropriate, we first identify the positioning requirements of each CV application listed in the USDOT?s Connected Vehicle Reference Implementation Architecture (CVRIA). These requirements include accuracy, integrity, update rate, and type of positioning (relative or absolute). Based on our overall analysis, we recommend two general positioning strategies: 1) utilize other sources of positioning information whenever possible (particularly at intersections), and 2) estimate the uncertainty of the positioning solution and use this uncertainty as an input to CV applications themselves.