Service assurance architecture in NFV

Service Assurance (SA) is a significant part of Network Function Virtualization (NFV) to enable automated and efficient service delivery from end to end (E2E). In NFV, SA should be integrated into the design and development loop from the beginning. However, it sees slower pace than other NFV management and orchestration (MANO) components. Most of present NFV-SA solutions are partial and do not provide a global view to assure services E2E. In this paper, we propose a NFV-SA architecture that on top of existing SA solutions, extends the state of the art by including an E2E scope and customer-facing assurance. The hierarchical, distributive and modular features of this architecture increases efficiency, reduces complexity and communications overhead. Furthermore, it gives operators flexibility to customize their SA framework and differentiate their services. We are also aware of challenges to realize this architecture. Learning from cloud computing, network analysis theory, and machine learning (ML), we propose a ML-enabled approach to address the challenges on monitoring and measurement.