There are two parts of the data: 1) a SUMO model of a typical intersection that consists of 4 approaches, each of which consists of 3 movements (left turn, right turn, and straight); 2) the vehicle trip information data generated by SUMO under different volume files, which is used to train the intersection signal control algorithm. The SUMO model contains five ".xml" files (node, edge, connection, net, and additional files) which are used to construct and configure the model. One can refer to the official SUMO tutorial for the format and functions of these files: (https://sumo.dlr.de/wiki/Tutorials/Hello_Sumo) The vehicle trip data is generated by SUMO as an output (which is specified in ".sumocfg" file). One can refer to the official tutorial (https://sumo.dlr.de/wiki/Simulation/Output/TripInfo) to understand the data format. Note that readers capable to read ".xml" files like Notepad++ are required to read the SUMO model and vehicle trip data.