Cooperative Localization using CARLA-SUMO simulators

This dataset is dedicated to the assesment of cooperative localization algorithms using realistic driving patterns from many vehicles moving in CARLA simulator, along with realistic V2V communication and network quality conditions.  t containts the trajectories of 60 vehicles moving for 3000 time instances in CARLA simulator, with driving parameters like 3D position, rotation, velocity, etc. Town01, which resembles a Manhattan grid topology, of the CARLA simulator is used to spawn a total of 60 vehicles. The synchronization step between the simulators was set to 0.1 sec. As for V2X communication, each vehicle had a transmitting power of 1mW, and the middleware update interval was configured to be 0.05 sec. Moreover, for each one of the agents we provide also the local dynamic map along the simulation horizon, as created by SUMO simulator, which contains the transmitted messages from neighboring vehicles. Those messages are comprised of neighboring vehicle id, message generation and expiration timestamp, 3D position, speed, yaw and yaw rate.