Multilayer Software Defined Networking Architecture for the Internet of Things

Abstract

The exponential growth in the devices connected to the internet of things (IoT) has raised lot of challenges for existing network architecture in providing support to time-constrained applications, device mobility, data management, etc. Fog computing enhances network performance by providing features like reduced response time, network and data security, etc. On the other hand, Software-defined networking (SDN), which slices out control and data planes, provides a platform for efficient network management. This strategy provides a flexible control mechanism for implementing policies of data and device management. In this paper, a new IoT architecture has been proposed which combines SDN features with fog computing so as to enhance the overall performance and management of network systems. The proposed framework also introduces an additional layer within the cell, which provides services to the IoT devices, thus reducing the latency and performs load balancing in the network. The new proposed IoT network architecture has been simulated using MININET simulator for evaluation of round-trip time (RTT) of each layer. The simulation results show a remarkable improvement in the latency of the proposed IoT architecture.