A Hybrid Systems Approach to Modeling Real-Time Situation-Awareness Component of Networked Crash Avoidance Systems

A new approach is introduced for modeling the estimation and networking processes of networked crash/collision avoidance systems (NCAS) in a single framework. An NCAS relies on real-time situation awareness, which is acquired through cooperation with neighboring vehicles in a wireless network, to predict and respond to hazards. The subsystem that provides situation awareness consists of an estimation process and a networking/communication process. The estimation process is the source of vehicle data that is broadcast through the underlying vehicular network (communication process). The performance of vehicle tracking (estimation process) is significantly affected by the performance of the communication network because any issues in sending or receiving the information could result in erroneous position estimates and possibly crashes. It is therefore essential to produce models that allow a clear view into the dynamics of these two components. We employ probabilistic timed automata to model the networking component; a hybrid automata is then used to combine and model both networking and estimation components in a single framework. We extend the presented hybrid automata to model the broadcast network of NCAS in presence of heavy hidden node interference. We verify model accuracy by comparison with proven network simulator 3 (NS-3) simulation models; model checking for the purpose of studying some features of NCAS is also presented.