A controller-based roadside unit plane architecture for software-defined internet of vehicles

The rapid development of the vehicle network has brought new challenges to the management for the internet of vehicles (IoV). As a combination of the IoV and software-defined networks (SDN), the software-defined internet of vehicles can provide an efficient network management method and effectively cope with the heterogeneity and high mobility of the IoV. Although the use of a distributed control plane to build a software-defined internet of vehicles is more suitable for a distributed architecture of the IoV, for a time-sensitive network such as the internet of vehicles, the choice of edge control plane also needs to consider factors such as flexibility and reliability. Delay optimization caused by controller plane improvement, deployment cost and possible management overhead is all problems that need to be paid attention to in software-defined internet of vehicles deployment. In view of the above problems, this paper selects the roadside unit (RSU) plane to construct the edge control plane of the software-defined internet of vehicles. The initial deployment of the control plane is first achieved by defining the RSU controller priority. The designed model mainly includes three factors: node contact time, flow scale and node popularity. Secondly, frequent fluctuations of the control plane are not conducive to network management, so the placement cost function and the RSU plane fluctuation cost function are designed to trade off the control plane dynamic adjustment overhead, thereby realizing the dynamic management of the control plane. Simulation results demonstrate that the proposed scheme outperforms the comparison schemes in four aspects: flow setup delay, end-to-end delay, number of controllers, and management overhead. Under normal operation, the proposed scheme is 20.3, 6.6, 34.1 and 4.9% lower than the best scheme IMABC in four parameters respectively. When the path failure probability p^fail = 0.02 exists, the proposed scheme is 19, 7.9, 28.6 and 2.7% lower than the best scheme IMABC in the four parameters, respectively.