Intelligent Control in SDN/NFV-Empowered IoT Systems for Smart City Application

Abstract

Artificial intelligence (AI)-based control approaches for management and orchestration layer are studied in this paper by investigating the applicability of emerging machine learning (ML)/deep learning (DL) modules in software defined networking (SDN) and network functions virtualization (NFV) control systems. The system architecture jointly integrates SDN/NFV empowered Internet of Things (IoT) for smart city applications. This study covers the coarse grained working flows of networking state observation, action configuration, and output metric evaluation in ML/DL-assisted modelling to obtain reliable performance metrics. The comparison between the conventional, SDN/NFV empowered IoT, and extended ML/DL-assisted systems are discussed to illustrate the performance improvement in terms of Quality of Service (QoS) metrics in the intelligent core networks. The simulation tools, such as Mininet, RYU SDN controllers (with FlowManager), and Iperf3, are discussed to deploy the IoT infrastructure topology, configure AI-assisted actions, and capture QoS metrics, respectively, to assist the planning before deployment