Analytical Model for Message Propagation in Opportunistic and Delay Tolerant Vehicular Ad-hoc Networks - 09/26/2008

Ashish Agarwal

The development of Dedicated Short Range Communication protocol (DSRC) is aimed at neighboring vehicles to exchange safety information to enhance awareness of the vehicle beyond the line of sight and enhance safety features such as active braking and collision warning. While safety is the primary goal for enabling inter-vehicle communication, several applications can be instantiated on the underlying infrastructure.

We consider the problem of inter-vehicle communication in a highway scenario where vehicles are partitioned, while there is opportunistic connectivity offered by mobility of vehicles. Thus, the partitioning is time varying, described as fragmentation of the network. Further, we assume a delay tolerant communication paradigm, which is essentially a store-carry-and-forward approach where nodes (vehicles) store data when the network is disconnected and adopt a greedy forwarding approach once connectivity is available. In this work, we model the delay tolerant network with opportunistic connectivity for parameters such as transmission range, node density and vehicle mobility. Our goal is to estimate bounds on the performance achievable for given parameters of the network.

Our results show that the performance of the network lies primarily in two regimes considering varying vehicular traffic density. One with low traffic density where the network is mostly disconnected and the other high density scenario where the network is fully connected. However, more interesting is the case of intermediate density case where gains are provided by intermittent connectivity and delay tolerant networking assumption. The results are presented in the context of the phase transition phenomenon observed. Furthermore, the conditions for gains achieved by the delay tolerant networking assumption and the those achieved by increased mobility of vehicles are presented.

In this talk, I discuss the analytical model in detail and present the results obtained from the on-going work.