Cluster Routing-Based Data Packet Backhaul Prediction Method in Vehicular Named Data Networking

What is it about?

Vehicular named data networking (V-NDN) is a network architecture that combines named data networking (NDN) and vehicular ad hoc networks (VANETs). Due to the high-speed mobility of the on-board unit (OBU) in V-NDNs, topological changes may cause the problem of reverse path breaking for data packets, thus impacting the communication quality of service (QoS) among vehicles. To address this issue, a data packet backhaul prediction method (DBPM) based on cluster routing in the V-NDN is proposed in this paper. The DBPM uses GPS and a convex programming location algorithm (CPLA) at roadside units (RSUs) to obtain the positioning information of vehicle in the clusters, and uses two positioning data items to predict the location of the vehicle’s future access point (AP) for the cluster by using the Kalman filtering model. Then, the DBPM forwards the returned data packets to the vehicle by the cluster. Simulation experiments are performed by using the simulators Simulation of Urban Mobility (SUMO) and VanetMobiSim. Results show that the proposed DBPM can effectively reduce the average delay and packet loss ratio in the vehicle-to-infrastructure (V2I) communication in urban scenes, thus enhancing the robustness of data transmission and effectively supporting the data communication’s QoS of V-NDN.

Featured Image

Photo by Hyundai Motor Group on Unsplash

Photo by Hyundai Motor Group on Unsplash

Why is it important?

Aiming at relieving the problem of the reverse path breaking for data packets, which is caused by the high-speed mobility of OBUs to support better QoS in the V-NDN, a cluster routing-based data packet backhaul prediction method was proposed in this paper. This method establishes cluster routing based on the clustering structure, reduces the average hop count of the data packets, and enhances the inter-cluster handover performance by the prediction of the target RSU. At the same time, the CPLA based on bisection sensing was improved to reduce the number of flooding broadcasts in the sensing process, enhance the positioning accuracy, and improve the prediction accuracy by using a combination of two prediction schemes.

Perspectives