Study On Routing Strategies In Urban Vehicular Networks

Routing in the urban vehicular network, which is a typical form of delay tolerant network, has always been a difficult problem. Mathematically, the non-deterministic vehicular mobility has introduced a great difficulty in the modeling of vehicular networks, which results in an unknown potential of performance of routing strategies in vehicular networks. Simulated results are relatively incomparable because of the heterogeneous methods adopted by different researches. This paper, therefore, addresses two problems: on the one hand, we proposed a theoretical optimal solution for routing in vehicular networks. This optimal routing algorithm has theoretically best performance in terms of end-to-end delay and hops. The performance upper bound can then be revealed; On the other hand, although the optimal routing algorithm is not a practical routing solution, we observed, studied the optimal routing traces and designed a practical routing strategy for vehicular networks by mimicking the key behavior of optimal routing algorithm. Our simulation was carried out on a platform based on the two-year GPS data of 4000 taxis in Shanghai. Simulations showed that the performance of this new practical routing strategy, compared with other existing routing algorithms, has an improvement of 50%.