Conference Paper

Architectures and Challenges Towards Software Defined Cloud of Things (SDCoT)

Mansour A.
Mokhtar B.

Last decade witnessed a growth on developing of IoT architecture models to cope with various challenges and domains requirements. This leads to rapid increasing the number of the connected devices, which means large amount of data are needed to be stored and processed. Form this context, Cloud Computing is emerged as a solution to enable centrally unlimited storage of data and programs with the ability to access anytime from anywhere. The integration between IoT and cloud brings the concept of Cloud of Things (CoT) to support different types of data and multiple services with data ubiquity, flexibility, security and resilience. CoT facilitates large-scale IoT heterogeneous networks interconnections, but it is valid for static infrastructures. Thus, new challenge opens to handle dynamic configurations and heterogeneous infrastructures over multiple geographical distributions areas. In this paper, we exploit the concept of Software Defined Network (SDN) to present a precursory description called Software Defined Cloud of Things (SDCoT). SDCoT opens new dimensions towards dynamic network resource managements and reconfiguration subject to QoS and performance requirements. This paper summarizes the proposed architectures for IoT and Cloud Computing with the corresponding challenges towards our initiatory attributive SDCoT. Though SDCoT, we can have a live deployment for different platforms and services with customized applications. This will facilitate implementation and configuration of sensors, gateways, switches or controllers based on the programed software defined services. This will rise a new Cloud definition called Device as a Service such as Sensor as a Services, Switch as a Service, Controller as a Service... etc. © 2018 IEEE.