Priority-Aware SFC Provisioning in Fog Computing

Fog computing paradigms are proposed to overcome delay limitations associated with cloud computing. This is accomplished by transferring the computational and storage resources, relaying and caching services from the cloud to the edge of the network. This in turn reduces delays and latencies and alleviates overloads at cloud nodes. Furthermore, virtualization technologies such as network function virtualization (NFV) facilitate accessing and sharing fog nodes by various services and applications. When using these technologies, service function chain (SFC) provisioning presents an important factor in providing such applications. Hence this paper presents a first study on SFC provisioning in NFV-based fog networks that takes into account shortest-path and load balancing to achieve efficient resources utilization. Batch requests are sorted based upon various priority-aware criteria, i.e., highest and lowest resources requirements for functions, as well as random sorting. Hence, the goal here is to achieve efficient SFC provisioning while identifying the best sorting technique for mapping batch requests onto the fog nodes.