Study of anti-congestion algorithms for autonomous and connected vehicles

Study of anti-congestion algorithms for autonomous and connected vehicles Modern vehicles are becoming smarter thanks to the continuous development of new Intelligent Transport Systems. For example, some new features ca be detect pedestrians on the road and automatically car comes to a stop avoiding a collision. Some others can be to obtain information about traffic congestion through the cellular network and suggestion to the driver of another route to save time. Nevertheless, drivers are always looking for a safer, cleaner, and more efficient way of travelling . In particular, the applied research industry is investigating on opportunities and benefits that can be brought by the introduction of autonomous vehicles (at whatever level of automation), because are considered safer than the guide of a human being. Accidents due to men are mostly caused by distracted driving, undecided performance and speeding. Bearing in mind the potential of autonomous and connected vehicles and the new needs that society faces, the thesis consists of the development of an algorithm, which is used in case of accident on highway with a fleet of connected and autonomous vehicles. It has the purposes: 1) to optimize flow and speed of cars on the lanes avoiding other accidents or slowdown and 2) to decrease the waiting time of the cars behind the obstacle without a sacrificial approach, when cellular vehicle-to infrastructure (C -V2I) is adopted as a communication technology. The vehicular mobility is obtained thanks to the traffic simulator Simulation of Urban MObility (SUMO). The infrastructure and the vehicles are represented in the application level of the open source OMNeT ++ network simulator, through the simple modules client and server. In particular, it is considered an LTE -A network with an eNodeB (eNB) located at the center of the topology. The server, hosting the developed algorithm, is located inside multi-access edge computing (MEC), which is in the eNB. The LTE-A network runs thanks to simuLT E, simulator for LTE/LTE-Advanced cellular networks in Frequency Division Duplexing (FDD) , representing the Radio Access Network and Evolved Packet Core. It is based on OMNeT + + and exploits the INET framework. The vehicles send CAMs ( Cooperative Awareness Message) messages to the server which takes decision about the single car with the aim to improve the traffic conditions. To verify the goodness of the algorithm, the results obtained by its application are compared with those given by the application of ALINEA, a ramp metering algorithm, and with those obtained by studying the behavior of the simulators only, in the same traffic conditions. The results show, that in the case of algorithms application, the possibility of accidents and other slowdowns is significantly reduced.