Service Function Chain Placement in Edge Computing: A Topological Dependency-Informed Approach

The integration of Network Function Virtualization (NFV) and Mobile Edge Computing (MEC) allows for efficient, advanced network services via Service Function Chains (SFCs). SFCs are sequences of ordered network functions designed to provide specific network services. However, the placement of SFCs is critical, especially for latency-sensitive applications such as telemedicine, due to spatial proximity between service functions, their processing order, and limited edge resources. This paper addresses the multi-SFC placement problem in MEC-NFV networks, aiming to reduce both deploying cost and routing cost. We propose an innovative algorithm based on topological sort, called TD-NFP, to solve this problem. The experimental results show that the proposed TD-NFP approach outperforms other benchmarks.