Quantitative Modelling and Automated Analysis of a Cooperative Congestion Control Protocol for Dynamic Vehicular Ad-hoc Networks

Vehicular Ad-hoc Networks (VANETs) are critical, self-configurable wireless networks for communication among vehicles and between vehicles and roadside equipment. VANET communication protocols must ensure safe and efficient communication among all network nodes within the signal range, in the context of high node mobility and rapidly changing network topology. In this paper we carry out a formal analysis of a cooperative congestion control protocol for VANETs, providing the first use of formal verification incorporating more realistic traffic models. We use Markov Population Processes to formalise and quantify message generation, rescheduling, and transmission, and vehicles joining and leaving the signal range. We analyse Quality of Service metrics concerning communication reliability and delays as functions of various parameters such as mean message size, vehicle density, road topology and traffic conditions. The modelling and verification approach is based on a counter abstraction technique that permits a full exploration of all possible behaviours by stochastic model-checking.