Probabilistic Road-Aware Geocast In VANETs

Abstract

Geocast is a communication technique to disseminate information in specific geographic regions instead of node addresses. Traffic congestion, accidents, local hazards and digital content sharing are potential use cases of information sharing in VANETs. Recently, several approaches for geocast routing have been proposed to achieve high delivery ratios. These approaches consider a center point and radius to define the destination region also called geocast region. They focus only on routing scheme to enhance the delivery ratio and delays. However, these approaches do not consider the target region selection problem in the geocast routing. In this paper, we propose a novel application-level mechanism for sharing road conditions, such as accidents, detours and congestion in VANETs through probabilistic road-aware geocast routing. We assign probabilities to the roads around each intersection in the neighborhood road network of the source vehicle. We then build a spanning tree of roads (from graph representation of the road network) with information source as the root node. Nodes below the root represent junctions and edges represent inter-connecting road segments. Messages propagate along the branches of the spanning tree. The spanning tree represents the geocast region. As the information propagates down the branches, probability of road as geocast region decreases. Information is propagated until a threshold probability is reached. Our method also ensures that messages are not delivered to irrelevant vehicles irrespective of their proximity to the source. We evaluate our application through extensive and realistic simulations in ns-3 simulator using IDM car following and MOBIL lane change models for realistic modeling of vehicle mobility.