A Solving Method for Computing and Network Resource Minimization Problem in Service Function Chain Against Multiple VNF Failures

Service Function Chain (SFC) has been an inevitable element for the development of Network Function Virtualization (NFV). SFC usually consists of some Virtual Network Functions (VNFs) such as firewall and network address translation (NAT). Although SFC inherently has scalability and flexibility, the routing for composing SFC and the VNF deployment tend to consume computing and network resources. Besides, multiple and simultaneous VNF failures inflict heavy damage on SFCs. This paper formulates a minimization problem of computing resource usage and a network resource usage for SFC against multiple VNF failures. Our proposed methods of SFC routing and VNF deployment aim to minimize the cost specified with the computing and network resources. Specifically, our SFC routing algorithm employs the k-node disjoint shortest path, and our VNF deployment algorithm employs the generalized assignment problem (GAP) and the set-cover problem. The paper also mentions intractability on the minimization problem and demonstrates the effectiveness of our solving method by conducting comparison experiments.