An NFV MANO Architecture with a Resource Allocation Mechanism Based on Game Theory

Abstract

This work presents an NFV management and orchestration architecture design to exploit resource allocation problems. We divide the architecture into six different modules: NFV orchestrator, VNF manager, infrastructure manager, VNF discover, VNF Monitor, and resource allocation. The last three ones are our main contribution. Specifically to resource allocation, we propose a game theory algorithm that aims to optimize the matching of the users' requests with Infrastructure providers' offers. In order to evaluate the proposal, we consider two different approaches to compare with our solution in terms of the number of resources allocated to users, the number of resources allocated by the infrastructure providers, and the final price of the transaction. Our solution was able to allocate around 30% more resources and save around 32% when compared with the other solutions in simulated scenarios.