Smart Grid Oriented Service Function Chain Deployment Method

The current smart grid service types show diverse characteristics and require dynamic allocation of network resources. To further improve the service quality and resource allocation rationality of power services, a resource allocation scheme for smart grid services is proposed to dynamically deploy virtual network functions to cope with the flexibility of upper layer applications. The service function chain deployment problem is modeled as a Markov decision process that can dynamically respond to frequently arriving online network function requests; reinforcement learning algorithms are introduced to optimize the deployment strategy. Experimental simulation results show that the resource allocation scheme has good results in handling NFV requests for smart grid services, significantly improving the service request acceptance rate and improving the benefit-cost ratio.