A hierarchical approach for accelerating IoT data management process based on SDN principles

摘要

The human life is changing drastically due to a popular emerging technology called Internet of things (IoT). The growth in the number of devices connecting to IoT networks results in large volumes of data collected or generated by these devices. Since IoT data is different from other types of data in terms of diversity and rapid growth and due to the huge amount of the generated data, the complexity in management and load balancing issues in IoT networks, there is a need for innovative ideas in IoT design and management. Software Defined Networking (SDN) as a promising paradigm separates the control functions from physical devices and places them inside a logically centralized control plane. SDN propounds the concept of programmable networks and due to its unique capabilities, seems to be very useful in reducing the management complexities of a dynamic network like IoT. In this paper, a hierarchical SDN-based approach is proposed for accelerating the data management and balancing the load not only between the IoT devices of a single domain, but also between different network clusters. The proposed framework prevents the controller from becoming a single point of failure and makes it possible to apply different kinds of management and load balancing strategies in a hierarchical and multistep scenario. According to the experimental results, our scheme exhibits a reduction in the average turnaround time and the average waiting time and improves the processing performance. Furthermore, the proposed approach distributes tasks evenly throughout the network and improves the use of network resources.