Four publications addressing recent topics in automotive communication networks presented at IEEE VTC-Fall 2019

Key results of the vehicular research projects "SFB 876 (B4): Analysis and Communication for Dynamic Traffic Prognosis" and "Intelligente Verkehrsinfrastruktur für sicheres vernetztes Fahren in der Megacity" (InVerSiv) were presented by CNI researchers at the IEEE VTC-Fall 2019. The IEEE Vehicular Technonology Conference (VTC) is the flagship conference of the Vehicular Technology Society (VTS) of the IEEE and took place in Honolulu during the 22nd-25th september. Two of our contributions focus on using machine learning methods to analyze and optimize cellular vehicular communication systems. In a comprehensive real world performance analysis, we analyzed online data rate prediction in different scenarios and networks of multiple operators. Based on the foundation of the obtained results, we demonstrated Data-driven Network Simulation (DDNS) as a novel approach for accurate and fast performance optimization of opportunistic vehicular transmission schemes. Hereby, a machine learning-based black box approach is applied for modeling the end-to-end behavior of a to be optimized network performance indicator. Futhermore, we made two open source contributions to recent topics in network simulation. Lightweight ICT-centric Mobility Simulation (LIMOSim) provides an integrated framework for the joint simulation of aerial- and ground-based communication systems and targets the integration of Unmanned Aerial Vehicles (UAVs) into vehicular traffic systems. In addition, we presented an open source simulation framework for the performance analysis of Cellular Vehicle-to-Everything (V2X) Mode 4 communication which is a controversially discussed candidate system for enabling connected cars.