Experimental investigation of mechanisms for congestion control in ITS-G5 based vehicular communication systems

Transport is fundamental to economic growth and quality of life. Information and communication technologies facilitate the exchange of data or information among vehicles and with the infrastructure. They are commonly regarded as a key to improve the future transport systems. For vehicular communication (Car-2-X communication), a WLAN-based system has been developed and standardized. The European variant of the WLAN for vehicular communication, ITS-G5, is, similar to IEEE 802.11p, derived from the IEEE 802.11a standard. Other important communication protocols are the GeoNetworking protocol and the Basic Transport Protocol (BTP) for ad-hoc communications, and application-oriented message formats such as the Cooperative Awareness Message (CAM). A major challenge for Car-2-X communication is to mitigate saturation and unlimited medium access delay caused by a potentially large number of vehicles. Decentralized congestion control (DCC) is a key functionality to solve this problem. The present diploma thesis studies DCC algorithms in an experimental setting. It implements a common framework for protocols of the facilities layer, integrates three candidate DCC algorithms in the framework, and carries out an experimental evaluation under various test scenarios. The experimental results show that the two DCC algorithms LIMERIC and resource distribution show a similar behavior for a specific set of parameters. The third DCC algorithm, state-based DCC, has a high delay in adapting the fast-changing channel load conditions, whereas the other algorithms show a fast adaption. Further experiments should take into account more message types from the facilities layer, use a variable size of CAMs, and extend the DCC algorithms with resource allocation mechanism.