Efficient Resource Management for Intervehicle Safety Communication

Safety systems based on intervehicle communication aim at reducing fatalities and injuries in road traffic. Vehicles are enabled to exchange information on their status. Active safety for driver and passengers is meant to be improved through reliable, low delay communication of highly accurate information, in all the various road traffic conditions. Currently in Europe, standardization regarding Intelligent Transportation System (ITS) is carried out by European Telecommunications Standards Institute (ETSI) For the application protocol layer, ETSI foresees three message types for broadcast: Periodic Cooperative Awareness Messages (CAMs), event-driven Decentralized Environmental Notification Messages (DENMs) and Service Announcement Messages (SAMs). Especially the communication load generated by the periodic CAMs may easily lead to an overload of the communication channel. This is a known issue of scalability in medium access in wireless networks with many nodes (>> 100 nodes) [21]. Common approaches proposed in the literature control single system parameters and aim at limiting the load to a certain threshold in order to avoid overlapping transmissions and thus prevent the well-known hidden station problem. This threshold is commonly set to a relatively low value, which leaves some resources unused. In this Ph.D. thesis, a more fine-grained exploitation of various system parameters is conducted. The author explores the optimization space and outlines important parameter limitations, which is confirmed by real-world experiments. As an outcome, a hybrid approach of single-parameter adaptation and cross-layer coordination is developed. The CAM generation rate is adapted to reduce the channel load while maintaining an acceptable position error. Forwarding of CAMs is employed selectively to repeat periodic messages with high importance. To manage the medium access under high load appropriately, the spatial reuse of the communication channel is adapted by adjusting the CCA threshold, especially for the safety-relevant area. The cross-layer management coordinates the single-layer mechanisms at application layer, facilities layer, and access layer. This coordination depends on the current channel condition as well as the road traffic situation. This so-called VANET Resource Management (VRM) considers requirements from different layers of the communication system as well as their inter-dependencies. To achieve manageable complexity by a lightweight design, VRM follows the model of a Finite State Transducer (FST). Instead of optimizing all parameters at the same time, VRM coordinates the singlelayer mechanisms which control small sets of input/output parameters. Thus, each approach has manageable complexity. Suitable examples for the application of the framework are proposed, evaluated and compared to existing approaches. The evaluation covers basic analyses of the constraints in wireless networks as well as road traffic. Real-world measurements underline the importance of an improved spatial reuse. Extensive simulation studies in large-scale scenarios are performed, covering isolated urban and high-speed highway traffic as well as mixed scenarios. Amongst others, it can be shown that VRM can achieve a more beneficial resource usage such that the awareness in the safety-critical area can be significantly improved.