JISE

The popularity of Network Function Virtualization (NFV) increases due to offering flexible and efficient methods for placing the chains of Virtual Network Functions (VNFs). The majority of recent studies consider consistent (immutable) bandwidth and flow pro-cessing while implementing the NFV-enabled network. The considering the consistent flow processing and bandwidth requirement become crucial in NFV-enabled network, where the network behaviour is highly dynamic. Due to the regular fluctuation in network traffic, the bandwidth demand, flow processing requirements, and the available Point-of-Presence (PoP) in the network change continuously. This article presents an adaptive model for the placement of VNF chains in order to mitigate the limitations mentioned above. The objective of re-arranging the prior placement of service chain is to maintain the performance of flow processing despite the variation in flow processing and resource requirements. Meanwhile, the proposed adaptive model minimizes the number of changes (i.e., re-allocation of VNFs) in the network. The simulation results show that the proposed model delivers the network services without abruption. Moreover, it also maintains the performance of flow processing and minimizes the 12% to 25% modifications in the pre-viously placed VNFs.