Real-Time Data Transport Scheduling for Edge/Cloud-Based Internet of Things

The rapid development of numerous smart-world applications powered by the Internet of Things (IoT) calls for innovation in network architecture to meet the increasing demands and service. In this paper, we first discuss the challenges of integrating edge/cloud computing infrastructures to support human network-of-things applications and propose a new three-tiered architecture that consists of IoT Things, Edge, and Cloud. We then design a Real-time Data Transport Scheduling (RDTS) scheme for efficiently transporting the data from IoT devices to cloud systems. The RDTS proactively allocates link bandwidth resources at the IoT Things and Edge tiers by assigning proper weights such that the network capacity can be effectively distributed to all devices. To demonstrate the effectiveness of the scheduling scheme, we conduct a performance evaluation. Our results show the RDTS scheme can significantly improve network performance with respect to throughput, delay, and energy consumption. Additionally, we discuss some remaining issues concerning the architecture.