Prediction-Based Geographic Routing over VANETs

The topology formed by vehicles changes quickly, which makes routing become instable. Geographic routing such as GPSR, compared with traditional routing, is more scalable and feasible. However, the commonly used greedy forwarding in geographic routing often fails, due to the urban topology in VANET is particular. Some improvements have been proposed, such as GROOV, which calculates the feasibility of every neighbor node to make the forwarding decision. And it depends on intersection coordinator to choose the next road segment that packets should relay to. However, it increases the packet travel hops when improves the packets delivery ratio. And it also forces packets to stop at an intersection, thus causes more communication traffic at intersections, which may bring congestion and possibly more relay hops. This paper proposes an enhanced geographic routing protocol called PGR, which enhances GROOV’s forwarding strategy and, through kinds of predictions, reduces the packet travel hops while maintains a relatively high packets delivery ratio. Simulation results show that PGR achieves a high level of routing performance in terms of hop counts, network latency and packet delivery ratio both in dense or sparse vehicular ad-hoc networks.