AODV improvement for vehicular networks with cross layer technique and mobility prediction

Vehicular Ad-hoc Network (VANET) is the road vehicle's communication network without a base station. For end-to-end communication between vehicles, a routing protocol is used to find a route based on link properties from one to another vehicle with specific requirements. In VANET, vehicles move at high speed, which makes topology changed all the time, often causing failures in data transfer. Additionally, the failures happen when the signal congestion is high in the VANET; the efficiency of data transfer therefore decreases. Thus, routing in VANET with high speed movement and high signal congestion is challenging. In this article, we present AODV-CAVM algorithm which including two mechanisms; first is using a cross-layer technique that calculates Channel Availability at the link layer; second is implementing mobility prediction based on neighborhood knowledge. We apply these mechanisms to modify AODV routing protocol to improve the communication reliability and latency for use in vehicle safety applications. We conduct our experiments on vehicular networks using NCTUns simulator with the realistic road-based mobility model. Our experiment results show that AODV-CAVM achieves better performance.