Cooperative driving plans and controls Connected and Automated Vehicles (CAVs) to move smoothly in conflicting areas, so that the traffic safety and efficiency could be significantly boosted. As a backbone of cooperative driving, road-side units receive massive data from the encountered vehicles via wireless communication and schedule the movements of vehicles based on these timely received data. This paper proposed a CD-DB storage model to achieve high write/read performance for supporting cooperative driving applications. We analyze the data structure and algorithms dedicated for cooperative driving applications. Testing results indicate that the proposed data storage model can handle real cooperative driving scenarios and yield significantly better performance than currently widely used relation databases.