With the softwarization of mobile wireless networks comes automated control and management of network infrastructure. Machine learning solutions come as critical enablers to achieve efficient network control and management. However, these machine learning solutions need data to train. In some applications, as is the resource allocation in the edge, large datasets, including User Equipment (UE) mobility between cells and traffic activity, are required. These may be difficult to obtain due to privacy concerns. This work presents a synthetic dataset generator that aims at supporting research activities in these areas. The introduced dataset generator uses traces from a known urban mobility simulator, Simulation of Urban MObility (SUMO). It matches them with empirical radio signal quality and diverse traffic models to obtain large datasets that can validate machine learning solutions. From the introduced generator, we created a dataset in an urban scenario in the city of Berlin with more than 6h of duration, containing more than 40000 UEs served by 21 cells.