Transportation Cyber Physical Systems (CPS) are expected to rely on robust wireless communication networks for real-time feedback for controlling these systems. The IEEE 802.11p based Dedicated Short Range Communication (DSRC) standard has been proposed for vehicular communications that has 7 channels. However, these channels could be easily congested resulting in delay and unreliable communications when vehicle density is high. In this paper, we present a cloud-assisted global positioning system (GPS)-driven dynamic spectrum access framework for transportation CPS. To provide reliable communications, we assume that each vehicle is equipped with two transceivers: one transceiver (always connected to the internet using e.g., 4G link) queries spectrum database and/or can serve as a GPS through an application (app), and the other transceiver/radio switches channels and adapts to suitable transmit parameters for vehicular communications to avoid any harmful interference to primary users (PUs). Each vehicle calculates the best route to its destination using GPS and finds the set of idle channels along the route. Furthermore, each vehicle periodically checks the spectrum database throughout the route to get most updated spectrum opportunities. We present performance evaluation of the proposed approach with the help numerical results obtained from simulations.