| Wireless communication between vehicles can greatly improve traffic safety andimprove traffic conditions. In VANET, the majority of the vehicles move regularly. Ina certain time, the moving trace is predictable. If two nodes contacts with each otherfrequently, they can be expected with higher probability contacts again. Now therouting protocol research about intervals of contact is mainly about two selectedvehicles. Because of the uncertainty of the vehicle moving trace, especially in thesparse nodes area, the interval will be very large.According to the feature of intervals, this article proposed a VANET routingprotocol based on the aggregate statistics about the intervals of contacts. Thealgorithm records intervals of contact between two select vehicles and aggregateintervals of contact in the vehicles flow. The protocol takes into account both thecharacteristics of the vehicle movement and the diversity of vehicle movingenvironment. In the data analysis process, we prove that the intervals distribution isregularly at first. Then we use the second-order Markov chain based on the historicaltrace to predict the next interval. In the forwarding process, we use the intervalsbetween the encounter vehicle and the destination to determine whether the encounternode will contact with destination faster than the current node. Then we calculate theaggregate intervals to determine the density of nodes in the moving environment. Ifthe encounter node can contact with destination in a shorter time or it moves in thearea with higher node density, the encounter node will be the next hop.Finally, we use QualNet simulation platform to evaluate the performance of thisprotocol. Compared with Epidemic protocol and the protocol only considering theinterval of two vehicles, we find that the protocol we proposed presented has arelatively stable performance in delivery ratio and average delay. Especially in thearea with sparse nodes, the average delay is far superior to other protocols. |
PDF Full Text Request