An End-to-End Security Model of Inter-Domain Communication in Network Function Virtualization
Håkon Gunleifsen, Thomas Kemmerich, Slobodan V. Petrović · 2016
This paper presents a new end-to-end security model for interconnected Virtual Network domains. Network Function Virtualization (NFV) has gained wide attention among Internet Service Providers during the last years. The standardization work from ETSI has outlined a common framework for Network Function Virtualization, open for multiple combinations of interdomain communication. The communication methods consist of multiple NFV interconnection technologies and interfaces, that open up for a variety of NFV models and increased complexity. From an Internet Service Provider (ISP) perspective, the ultimate goal is to be able to freely interconnect NFV services with other ISPs in a secure and automated manner. Hence, this paper presents an abstraction model of the current NFV end-to-end network transport mechanisms for inter-domain communication, to model the end-toend security. The general work within the NFV domain is driven by multiple research contributors where academia, standardization organizations and the open-source community further develop the technology. To verify the model and contribute avoiding research silos, it is also important to classify the related research. We use the presented model for such classification of NFV interconnection mechanisms. By categorizing the differences between the NFV interconnection layers, we show that the model can be used to identify the security gap for secure network channels in NFV.