An efficient load balancing approach for service function chain mapping

Abstract

Network function virtualization promises a significant advantage for addressing diverse quality of service requirements via chains of Virtual Network Functions (VNFs). Service Function Chain (SFC) mapping, composed of placing VNF in the physical network and routing traffic through them, is the key problem to meet the requirements. In the recent mission-critical applications, delay guarantee and fault tolerance are the crucial requirements wherein links and servers load balancing greatly facilitate providing the requirements. In this paper, we approach the problem of minimizing the maximum load of links and nodes in SFC mapping. After formulating the problem as a mixed-integer linear program, a batched version of the model is developed to estimate the optimal solution in a reasonable time. Then, we propose a practical solution based on the water-filling algorithm. Simulation results in different scenarios show that the algorithm can yield excellent performance compared to the optimal solution and benchmarks algorithms.