Cross-Layer Greedy-Based Routing in VANETs: Virtual Multibackbones Approach

Abstract

An efficient data packet delivery in a vehicular ad hoc network (VANET) is still a challenging task because of the rapid changes in the network topology and the instability of link quality, especially for nonsafety applications. Several routing schemes are proposed to improve the reliability of data delivery between vehicles. These protocols still suffer from the growth of the data packet delivery delay, the lost data packets, and the control packets. For this reason, we propose a novel cross-layer routing protocol, which is the first that uses virtual multibackbones and combines them with the quality of the link to choose the next forwarder vehicle. The next candidate forwarders, among neighbors, are those that keep a minimum of link quality during the data packet delivery delay. Then, among these candidates, we select that is closest to the destination vehicle to receive and forward the data. In addition, our proposal gives priority to those that travel toward the destination vehicle avoiding the forwarding of data several times by the same vehicle. The substantial simulations show that our protocol outperforms certain of the well-known existing schemes in the VANET area by varying the vehicle density on the highway. It exhibits an improvement of 6.72% in the average end-to-end delay of data delivery, 14.44% in the packet delivery ratio, 11.2% in the throughput, and 6.4% in the control overhead as compared to other schemes.