An Energy-Efficient Controller Management Scheme for Software-Defined Vehicular Networks

Abstract

The next generation of sustainable vehicular networks is expected to have a wide range of access technologies, multi-homing capabilities, and traffic demand and pattern heterogeneity. In order to deliver massive data loads from various services and applications to end-users, network softwarization is a key contributor that mitigates challenges in heterogeneous networks by providing a shared interface platform. SDN-enabled vehicular networks provides global bird's eye view of the network's status and connectivity. However, using a centralized control center, brings many difficulties, including bottlenecks and densification issues. A distributed control plane is an alternative, but it raises questions about where to deploy the control units and how many controllers are needed in a given network structure. In this article, we propose an energy-efficient adaptive controller management strategy for distributed software-defined vehicular networks using vehicles' mobility densities and communication latencies between switch-enabled access points. To reduce the number of controllers, the proposed method employs a split-and-merge clustering technique. The performance of the clustering solution was evaluated using realistic mobility traces and compared to a number of benchmark clustering solutions and variations. The results indicate the efficiency of the proposed scheme in terms of energy consumption, latency, and load on network entities.