Association-based optimal data forwarding in vehicular networks

Vehicular ad hoc networks (VANET) have attracted considerable attentions in recent times for promising applications like vehicular safety, traffic monitoring, etc. Due to fast topological change and frequent network disruption, multihop data dissemination is an important issue. In this paper, we propose a greedy intersection-based optimal routing protocol, in which we formulate the data forwarding problem as a Markov Decision Process (MDP) with associated states. Our protocol is aimed at moving destinations whose trajectories are unknown beforehand. We utilize realistic Shanghai SUVnet mobility traces to conduct extensive simulations and the experimental results illustrate a better performance than previous works.