With the introduction of network function virtualization (NFV)

operators can provide more flexible network service. However

most existing orchestration schemes of service function chain (SFC) are limited to the static or reactive policy where the virtual network functions (VNFs) need to be deployed or migrated frequently

and will easily lead to the mismatch of service supplement and high operational expenditure under time-varying workload. This paper proposes an online elastic orchestration policy (OEOP) based on the evolution perception of flow rate to solve the above mentioned problem. OEOP introduces online learning into the evolution perception of flow rate

which helps to predetermine the fine-grained VNF scaling demands. In addition

the online elastic deployment is achieved according to the real-time update information of SFC paths and the load of nodes. The newly deployed VNF instances can respond to the time-varying workload by taking place of the mission of VNF migration. The simulation results demonstrate that OEOP can significantly enhance the matching between virtual resource supply and workload demand. The throughput of VNF is improved by 10.2%~24.8%

and the operational expenditure can be reduced by 26.7% on average compared with other solutions.