- Lower CAPEX by reutilizing dedicated appliances with shared servers
- Reduce operational costs with centralized pool of appliances to maintain
- Enable Innovation by making it easier to develop VNF
- Speed up revenue-generating services to market
- Support on-demand pay-as-you go deployment models
- Deploy virtualized solutions on commercial, off the shelf (COTS) hardware
- Take network functions in and out of service, and easily scale them up, down, and out
- Centralised data management for Operators having deployment on various sites
Network Function Virtualisation (NFV) is not a concept, which has come into development recently. It has been around since 2012 when Deutsche Telekom, China Mobile, BT, AT&T and many other Telcos introduced the concept of NFV. The goal of NFV is to separate hardware from applications to eliminate the dedicated hardware devices for network functions, such as routers, firewalls, load balancers etc to be hosted on Virtualized platforms through Hypervisor. The Hypervisor is a virtual machine manager that allows multiple operating systems to share a single hardware host. There are a lot of use-cases associated with NFV of which we would be discussing some.
NFV Infrastructure as a service
Operators find the cost of dedicated standalone applications exorbitant and even difficult to maintain. The concept of NFV helps in addressing the growing and ever changing customer requirements, while controlling the CAPEX & OPEX overheads. Operators may decide to go for coexisting architecture with virtualized and non-virtualized functions. CSPs can build their own NFV infrastructure or go for NFV infrastructure as a service from service providers.
With CSPs looking to expand their reach to address global customer requirements, installing Virtualized Network Function (VNF) instance on a NFV as a service platform on cloud, enables CSPs to extend their reach instantly, and meet dynamic customer requirements. The CSPs that develop NFV Infrastructure Platform, enable additional commercial service offer to other CSP for NFV as Service (NFVaaS)
Virtualization Mobile Core Network components
Successful delivery of applications from a data center is seen as an important evolution and competitive advantage for the Operator. Turning the manual process of provisioning and delivering applications into a simplified, automated, policy-driven process can yield significant cost savings, as well as improvements in security compliance for the Operator.
Network Function Virtualization of Mobile Core components enable reduction in complexity in heterogeneous networks. Flexibility in deploying high load applications / Software on common hardware and storage will lead to significant reduction on total cost of ownership and improve network usage efficiency in operations, also can easily help accommodate increased demand for a particular service.
For example- PCRF is a key element in EPC architecture and is required to function as an end to end system with other EPC components, such as Online charging system that does balance management and balance management for ongoing sessions.
Resource Pooling
Operators who have a global presence had their resources set-up at various locations in the past but with the development of NFV all the resources can be pooled together setup at a single location with a dedicated bandwidth between the locations. There would be an ‘orchestrator’ located at a central location which manages the on-demand scalability of the deployment with an EMSs aid. The EMS monitors the system KPI , on reaching predefined KPI threshold it can send trigger to the orchestrator which in turn adds or removes the VNFs accordingly.
Benefits of NFV Solution: