A Survey on the Placement of Virtual Network Functions

Jie Sun,Yi Zhang,Feng Liu,Huandong Wang,Xiaojian Xu,Yong Li
DOI: https://doi.org/10.1016/j.jnca.2022.103361
IF: 7.574
2022-01-01
Journal of Network and Computer Applications
Abstract:The dependence of traditional network functions (NFs) on special hardware results in high capital expenditures (CAPEX) and operating expenditures (OPEX). Network Function Virtualization (NFV) is considered a promising technology to reduce the specificity of network equipment as well as the CAPEX and OPEX. With the decoupling of software and hardware, the flexible placement of virtual network functions (VNFs) resolves a wide range of network issues ranging from delay to reliability, from cost to performance, and so on. In this paper, we conduct a comprehensive survey of the VNF placement problem (VNFPP). Firstly, we provide a generic definition of the VNFPP by proposing and combining the basic models of chaining, placing, embedding, and routing. Secondly, we construct a fine-grained taxonomy of works in the VNF placement field based on the solving steps, i.e., objectives, constraints, models, and algorithms. In addition to the traditional cloud network scenario, we investigate the research progress of VNF placement problems in emerging placing scenarios, e.g., backbone networks, mobile networks, and the Internet of Things. Finally, we point out the open challenges and research directions in the field of VNF placement. Compared with the existing surveys, our survey provides a more granular taxonomy and a more comprehensive analysis of the state-of-the-art works in the VNFPP field.
What problem does this paper attempt to address?