A Two-Stage Heuristic SFC Deployment Approach in Software Defined Satellite Networks

Integrating network function virtualization (NFV) into service function chaining (SFC) allows for the flexible processing and forwarding of traffic along predetermined virtual network functions (VNFs), thus effectively coordinating resource allocation in software-defined satellite networks (SDSNs). However, the deployment of SFC in satellite networks is more complex than in terrestrial networks due to certain characteristics of satellites, such as dynamic topology and limited payloads. This paper proposes a delay minimization problem that couples VNF deployment and routing, and then presents a two-stage heuristic algorithm to solve it. The first stage employs an improved particle swarm optimization algorithm, while the second stage is a traffic routing algorithm based on Time-Evolution Graph (TEG). Finally, simulations are conducted under various system settings to validate the effectiveness of the algorithm.