Link Quality-Aware Geographic Predictive Routing for V2V Network Based on GPSR

Abstract

In recent years, advances in automotive IoT technology have not only improved automated driving and driving convenience but have also led to a diminishing number of traffic accidents. Vehicle Ad-hoc Networks (VANETs) have attracted much attention because automotive IoT is based on wireless communications. Ad-hoc routing protocols that apply the characteristics of vehicle-to-vehicle (V2V) networks are needed. One of them is Greedy Perimeter Stateless Forwarding (GPSR), a routing protocol using geographic information. It uses Global Positioning System (GPS) information for routing. However, GPSR may increase the packet drop rate due to radio attenuation and fading. Therefore, C. Li et al. proposed to select the next hop by evaluating the link quality. However, due to the large moving speed of the node, the next hop may be out of the radio propagation range at the communication timing. Therefore, in this study, we propose a new routing method that combines the future location prediction proposed by Z.S. Houssaini et al. The effectiveness of the method is demonstrated by simulations using ns-3 with traffic data acquired by SUMO.