Reliable Service Function Graph Construction and Mapping Mechanism for Cloud-Edge Collaborative IoT

With the development of 5G and Internet of things (IoT), data generated from edge applications is growing rapidly. Since processing all the data in the cloud will cause high network load and propagation delay, an NFV-enabled cloudedge collaborative IoT (CECIoT) architecture is built to jointly utilize the computing capacity both at edge and cloud through service function chain (SFC) deployment. However, current SFC construction methods always sequentially arrange the virtual network functions (VNFs) and seldom consider to process the VNFs in parallel, leading to a relatively high delay. In addition, it is vital to balance the reliability and resource utilization during backup. To address the challenges, we propose an SFC mapping optimization model for IoT applications to improve the quality of service (QoS) and the resource consumption. Then, we design a reliable service function graph (SFG) construction and mapping mechanism, which parallelizes and combines the SFCs into a graph, as well as provides the multi-SFC-sharing backup. The simulation results show that our proposed mechanism performs better than comparison methods in terms of QoS and deployment cost while providing differentiated services for IoT applications.