Time Constrained Message Delivery Probability in VANETs

Vehicular Ad-hoc Networks (VANETs) have recently attracted a lot of attention, as a tool to disseminate information among vehicles with the dual purpose of increasing road safety and comfort in driving. The majority of messages exchanged by vehicles are characterized by a finite lifetime period, after which, their level of usefulness is greatly reduced. Thus, an important problem in VANETs is to pursue efficient routing policies which can ensure information delivery in target areas, before this deadline expires. The routing problem has been studied extensively in the literature and a number of solutions have been proposed. However, no previous work has investigated the problem based on the probability to deliver messages to a specific area in a certain amount of time, in a setting where no static infrastructure is used. In this paper we study the problem using a graph theoretical approach. We model the road map and the traffic characteristics using a directed weighted graph which we refer to as the Road-Graph. We use the Road-Graph to calculate the probability of disseminating information along any given path, in a specific amount of time. So, for any two points on the Road-Graph we can find the path with the maximum probability of successful information delivery in the chosen time interval. This information is of great significance since it can be utilized by routing protocols to optimally route packets in the vehicular network. We validate our analytical findings by comparing them with data obtained using the VISSIM simulator. The simulation model that we consider represents a section of the Highway network in the Los Angeles area and is tuned using real data. We observe good matching between the analytical and simulation results thus demonstrating the validity of our approach.