A scalable metaheuristic for service function chain composition

Network Functions Virtualization (NFV) is a new network paradigm that has been strongly promoted from both scientific community and telecom industry, where network functions (NFs) such as firewalls, load balancers, gateways among others, are virtualized, isolated from middleboxes and housed on one or more industry standard computing nodes. One of the main challenges for service providers when they try to deploy the NFV-based networks is to efficiently make use of substrate network resources to facilitate the design, delivery and operation of network services in a dynamic and scalable manner. This challenge is known as NFV Resource Allocation. Service Function Chain (SFC) composition is the first stage of the NFV Resource Allocation problem and we solve it in this paper. We formally define the service requests (VNFRs) in terms of NFs and propose a metaheuristic algorithm for solving the SFC composition stage. Our evaluation focuses primarily on minimizing the total bandwidth demanded by the constructed network service. The results show that our proposed approach is applicable to large and complex VNFRs and finds, in reasonable running times, close-to-optimal solutions with minimal bandwidth requirements.