<p>In the last decades, there has been a trend to virtualize computing and networking resources over the physical infrastructure. This approach was initially adopted in the core functions of the network, thus creating the network function virtualization paradigm. The process which defines the computational nodes and links to execute a set of virtual network functions (VNFs) is called placement. However, executing only VNFs individually is often not enough to meet users’ requirements. Therefore, the service function chain (SFC) concept was created. With the advancement of the edge-cloud continuum infrastructure, SFCs started to be executed on multiple nodes, sometimes managed by independent service providers. In this multi-domain scenario, a distributed placement mechanism to allocate SFCs without complete knowledge of the infrastructure where the VNF instances will be executed is required. In this paper, we propose a new approach, named SPEED, to solve the SFC placement problem (SFCPP) over a multi-domain environment in a distributed manner. The solution encompasses algorithms and a new mapping model between the SFCPP and the game theory approach. The results of a simulation-based performance evaluation show that SPEED is feasible and, compared to other approaches, has a gain of 30% in the number of SFC requests placed successfully.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Speed: a distributed approach for SFC placement in the edge-cloud continuum

  • Anselmo Luiz Éden Battisti,
  • Flavia C. Delicato,
  • Débora Christina Muchaluat-Saade

摘要

In the last decades, there has been a trend to virtualize computing and networking resources over the physical infrastructure. This approach was initially adopted in the core functions of the network, thus creating the network function virtualization paradigm. The process which defines the computational nodes and links to execute a set of virtual network functions (VNFs) is called placement. However, executing only VNFs individually is often not enough to meet users’ requirements. Therefore, the service function chain (SFC) concept was created. With the advancement of the edge-cloud continuum infrastructure, SFCs started to be executed on multiple nodes, sometimes managed by independent service providers. In this multi-domain scenario, a distributed placement mechanism to allocate SFCs without complete knowledge of the infrastructure where the VNF instances will be executed is required. In this paper, we propose a new approach, named SPEED, to solve the SFC placement problem (SFCPP) over a multi-domain environment in a distributed manner. The solution encompasses algorithms and a new mapping model between the SFCPP and the game theory approach. The results of a simulation-based performance evaluation show that SPEED is feasible and, compared to other approaches, has a gain of 30% in the number of SFC requests placed successfully.