What Is Network Function Virtualization (NFV)? Meaning, Working, and Importance
In network function virtualization, network functions powered by hardware are replaced with software.
Network function virtualization (NFV) is defined as a telecommunications networking solution where network components traditionally powered by hardware are replaced with software that fulfills the same functions. NFV-powered networks are easier to standardize and more flexible, expandable, manageable, and cost-effective. This article covers the meaning, working, and importance of NFV.
Table of Contents
What Is NFV?
Network function virtualization (NFV) is a telecommunications networking solution that replaces network components traditionally powered by hardware with software that fulfills the same functions. NFV-powered networks are easier to standardize and more flexible, expandable, manageable, and cost-effective.
Simply put, NFV virtualizes major network components, removing the need for dedicated networking hardware for different functions and replacing it with software installed on a server. This enables entire network node function classes to be configured as building blocks that can be interlinked to set up end-to-end telecommunications networks.
While NFV uses traditional server virtualization methodologies, they are extended significantly. This helps telecom operators replace custom hardware appliances for network functions with virtual machines (VM) powering various software programs and processes. These VMs run on high-volume servers and provide the same functionality as switches, databases, and cloud computing platforms.
Virtualized functions provided by NFV include load balancing, intrusion detection, firewalls, routers, WAN accelerators, billing, and access control.
How NFV was Developed
Developing any telecommunication networking solution requires fulfilling rigorous standards for stability, quality, and protocol adherence. However, when it comes to hardware development, adherence to these standards often leads to longer product cycles, slower development, and increased risk of vendor lock-in.
When traditional telecom was the dominant communication option, this was not seen as a hindrance. However, the emergence of swift online communications services increased the demand for more responsive networks, disrupting the status quo.
In October 2012, a white paper on OpenFlow and software-defined networking (SDN) was published by an NFV working group from the European Telecommunications Standards Institute (ETSI). This white paper gave impetus to the broader adoption of network function virtualization.
With more users now aware of the benefits of virtualization in networking, vendors began to focus on improving IT virtualization technology. This led to an increase in the availability, scalability, performance, and network management capabilities of NFV. As with other technologies, minimizing the total cost of ownership (TCO) has been critical for effectively implementing carrier-grade features.
Virtualization has revolutionized how administrators specify, implement, and measure network function availability. VNFs are increasingly replacing traditional networking equipment. Limited equipment availability is less of a challenge than before, thanks to a layered, comprehensive, service-based approach adopted by network function virtualization providers.
As NFV decreases reliance on specialized equipment for specific functions, availability thresholds increase, with redundant resources effectively leveraged to obtain as high as five-nines (99.999%) availability.
NFV technologies are also capable of virtualizing numerous function types. They support a wide range of fault tolerance methodologies. Additionally, a high level of flexibility allows NFV vendors to fulfill modern-day networking requirements.
How NFV is used
Cost-effectiveness is a crucial advantage of NFV. Network operators reduce costs and boost the speed of service deployment by unlinking vital network functions such as encryption and firewall from dedicated hardware and shifting them to virtual servers. This removes the need for installing costly and complex proprietary hardware and allows for purchasing economical switches, servers, and databases to operate virtual machines for various network functions.
By consolidating numerous network functions into fewer physical servers, costs are reduced, and maintenance and management are simplified. In case a new network function is required, the vendor can simply set up a new virtual machine to perform it instead of installing new hardware.
Unlike a virtualized network–which, as the name suggests, entails the virtualization of an entire network–NFV virtualizes only specific network functions. Let’s say a vendor needs to enable end-to-end network encryption. Traditionally, this would mean deploying new hardware across the network. However, with NFV, an encryption software solution is deployed instead, using a standardized server or switch already set up in the network. Reduced hardware dependency translates to enhanced cross-network customization and scalability.
See More: What Is Local Area Network (LAN)? Definition, Types, Architecture and Best Practices
How Does NFV Work?
The core purpose of network function virtualization is the deployment of virtualized components for networking, allowing users to set up hardware-independent infrastructure. Server virtualization technologies are used by NFV architecture to set up the virtual machines that support network operations hosting.
Businesses leverage virtualization to allocate on-demand resources for addressing the requirements of developing dynamic workloads. Another critical advantage of virtualization is the reduction of costs associated with commercial off-the-shelf (COTS) hardware, such as x86 servers, which are used for virtualizing and deploying everyday computing, network, and storage resources. NFV replaces software with hardware for load balancing, firewall security, and routing functions.
But how does all this happen? Operationally speaking, NFV engineers use hypervisors to program various virtual network components and automate network deployment. NFV also enables IT managers to modify several aspects of network operations swiftly by using a single control point.
Thanks to NFV, IT resources such as virtualized data centers are better utilized. Automated resource allocation means that each virtual machine is assigned a specific portion of the available resources on a single x86 server. This allows each server to host multiple VMs, which can scale up to use available resources as required. The data and control planes operate within the data center and on external networks.
NFV architecture
Open and flexible architecture is a crucial feature of network function virtualization that gives users access to numerous deployment choices.
A typical NFV architectural framework consists of three main layers:
- VNF
Virtual network functions (VNF) are the building blocks of the NFV architecture. A VNF is a virtualized network component, such as a virtual router, firewall, base station, DHCP server, or network sub-function. For instance, numerous sub-functions, including mobility management entity (MME), home subscriber server (HSS), and serving gateway (SGW), serve as independent VNFs and collectively function as a virtual evolved packet core (EPC). VNFs are deployed on VMs, with a single VNF being deployable on either a single VM or across multiple VMs. Each VM can host a single VNF function or a subset of the overall VNF functions in the latter.
An element management system is a subsection of VNF. It supports functional VNF management such as configuration, fault, performance, security management, and accounting. This system uses proprietary interfaces to control either a single VNF or multiple VNFs simultaneously. Element management systems can be deployed as VNFs too.
- NFVI
Network function virtualization infrastructure (NFVI) encompasses all the hardware and software elements used to create the framework for VNF deployment. Users access NFVI to manage and execute VNFs.
An NFVI setup can physically exist across several locations simultaneously, with the network providing connectivity to create a comprehensive framework. NFVI includes the hardware layer, the virtualization layer, and virtual resources.
The hardware layer of NFVI includes IT infrastructure such as computing, network, and storage elements. These elements provide VNFs with processing, connectivity, and storage functionality through the hypervisor. Storage and computing resources commonly exist in a resource pool, with the network resource comprising switching functions such as routers and wired and wireless networks.
The virtualization layer is synonymous with a hypervisor that functions by abstracting hardware resources and decoupling VNF software from its underlying hardware. This layer ensures that the VNF lifecycle remains hardware-independent. Its main functions include the abstraction and logical partitioning of physical resources, a process commonly occurring in the hardware abstraction layer.
The virtualization layer is also responsible for enabling the software-based implementation of VNF to grant access to the underlying virtualization infrastructure. Additionally, this layer provides virtualized resources to allow VNF execution. This layer allows VNFs and hardware resources to be independent, and software deployment on various distributed physical resources becomes possible.
Finally, virtual resources are formed when the virtualization layer completes the abstraction of the storage, computing, and network functions from the hardware layer and makes them available for allocation and use.
- NFV MANO
NFV management and network orchestration (NFV MANO) is the layer responsible for managing and orchestrating roles within the NFV architecture. Its functions include end-to-end resource management such as computing, storage, networking, and VM resources in virtualized data centers. The main goal of NFV MANO is to enable flexible onboarding, which helps manage the uncertainty associated with the quick spin-up of network elements.
This framework was developed by the NFV MANO working group associated with the European Telecommunications Standards Institute (ETSI) Industry Specification Group for NFV (ETSI ISG NFV). This framework became more commonly known as ‘NFV management and orchestration’ with time.
NFV MANO can be classified into the following functional blocks:
- The NFV orchestrator addresses onboarding for new VNF packages and network services, global resource management, NS lifecycle management, and the authorization and validation of NFVI resource requests.
- The VNF manager is responsible for the lifecycle management of VNF instances. This block undertakes the adaptation and coordination role for event reporting and configuration between NFVI and element/network management systems.
- The virtualized infrastructure manager (VIM) manages and controls the NFVI storage, network, and compute resources.
The effective operation of the NFV MANO architecture relies on integration with open application program interfaces (APIs) in already-deployed systems. The MANO component can operate with standard VNF templates and allows users to choose from current NFVI resources to deploy an element or platform.
An operator’s decoupled operation support subsystem (OSS)/business support system (BSS) layer can be integrated with NFV MANO through standard interfaces. OSS is responsible for managing networks, faults, configurations, and services. BSS addresses the management of customers, products, orders, and so on.
NFV implementation
Now that we’re familiar with the architecture of a network function virtualization setup let’s see how users can implement it.
The first step is creating and deploying VNFs, which must be strategically and sequentially built out as part of a service chain. This will support the delivery of complex products and services.
The orchestration process is another step for NFV implementation. The network orchestration layer is responsible for deploying, monitoring, repairing, and billing VNF instances. Implementing carrier-grade features allows for highly scalable and reliable services, ensures high availability and security, and helps minimize operational and maintenance costs.
It is critical for the efficiently implemented orchestration layer to manage VNFs independently from the underlying hardware. Any VNF running on any technology from any supplier must operate smoothly.
A key element of NFV equipment is reliable, high-performance servers. The NFV architecture leverages server virtualization technology, with OpenStack, VMware, and container technology being the prominent options for the virtualization layer. VMware and OpenStack also serve as the leading hypervisor options. Container-based NFV may have yet to see widespread adoption, but it offers cutting-edge benefits for application performance.
Architectures for the NFV MANO layer vary, with some being open standard and others being proprietary. One of the leading options for open-source MANO is the Open Network Automation Platform (ONAP) by the Linux Foundation. Network operators must customize NFV MANO according to the specific needs of their operations and billing architectures.
The application layer enables VNFs to provide full-featured network application code. More advanced configurations allow network operators to choose multiple VNFs for ‘service chaining’ to deliver extensive network functioning. The applications of NFV span numerous network functions. Typical NFV applications include:
- Mobile networks
- Content delivery networks (CDN) and content delivery services such as video streaming
- IP multimedia subsystem (IMS)
- Evolved packet core (EPC)
- Virtual customer premises equipment (vCPE)
- Network slicing
- Network monitoring
- Software-defined branch and SD-WAN
- Web application firewalls
- Load balancers
- Session border control (SBC)
- Cybersecurity functions such as intrusion detection and prevention systems, firewalls, and network address translation
NFV vs. SDN
The terms network function virtualization and software-defined networking (SDN) are often used interchangeably. However, this is not accurate; while the two solutions may be closely related, they are not the same.
SDN replaces standardized networking protocols with centralized management and control. Its goal is the reduction of the complications associated with the distributed control of networking protocols using the simplicity of programming a comprehensive controller.
Additionally, SDN separates the control and forwarding planes of the network, giving users a centralized view and allowing for the swift and efficient implementation and operations of network services. SDN significantly boosts flexibility as only one instance needs to be updated for a change to be reflected.
On the other hand, NFV replaces proprietary network elements with software operating on standard servers. Rather than centralizing control, NFV optimizes network services. It focuses on decoupling network functions from the underlying proprietary hardware and shifting them to generic computers. This enables network functions to embrace software-powered operations and provide higher operational flexibility. Changes and updates to the network are simplified thanks to NFV.
Granted, both solutions use software to make networking more efficient; however, their goals are different. The contrasting features between SDN and NFV allow both technologies to be used on a single network in a way that allows them to complement each other.
See More: What Is a Wide Area Network (WAN)? Definition, Types, Architecture and Best Practices
Benefits And Challenges Of NFV
Now that we are familiar with the workings of network function virtualization, we’ll take a closer look at its benefits and challenges:
Benefits of NFV
- Cost
As companies expand, they increasingly invest in high-cost, dedicated networking hardware such as firewalls, load balancers, and routers to provide reliable networks. Such dedicated hardware demands more power and space for efficient operations, further increasing the capital investment requirement.
Conversely, NFV allows users to deploy generic high-capacity servers along with virtual machines, which are relatively affordable, especially on a larger scale. One can further minimize networking costs with the help of cloud-powered network virtualization abilities, which enable NFV solutions to efficiently deploy software across network locations and minimize hardware infrastructure requirements.
- Flexibility and scalability
By facilitating seamless network service provisioning, NFV boosts network flexibility and helps operators address dynamic consumer demands effectively at scale. Users can deploy new services as required. Additionally, enterprise network systems powered by NFV feature higher adaptability due to simplified installation and provisioning processes.
- Security
Due to network security concerns, companies are increasingly seeking greater control over their networks’ management. NFV secures networks by implementing industry-standard virtualized security gateways for server environments. Further, NFV can secure enterprise networks by deploying virtualized solutions such as access control, encryption, anti-malware, and intrusion detection and prevention, making network security more cost-effective and agile.
- Operations
By aggregating company-wide network functions in software form, NFV simplifies the remote transfer of network components from one network location to another. Users no longer need to rely heavily on on-premise network functions. For instance, network engineers do not need to visit a site to install a hardware firewall; instead, it can be provisioned remotely.
Challenges of NFV
- While large-scale NFV deployments are more economical, they can also present occasional challenges regarding reliability.
- Difficulties in coexisting in cloud-powered hybrid frameworks alongside physical network devices.
- The need for abstract network management, which requires a comparatively more specialized skill set.
- Highly dynamic network environments that require greater responsiveness across other business facets.
- The need for process realignment in enterprises that upgrade existing networks using NFV, without which managing traditional and virtual infrastructure simultaneously can become difficult.
Meeting the challenges of NFV
A complete transition to NFV can only happen after a few days. Existing legacy networks will remain in place for quite some time–perhaps years–while business processes catch up.
Some ways in which enterprises can meet the challenges of NFV architecture are:
- Support for real-time and dynamic service and network changes to address specific network events.
- Network services are being supported using a modeling approach.
- Decoupling of network state management and network configuration.
- Support for SDN controllers and network orchestration platforms.
- Comprehensively planned migration strategies before the deployment of network virtualization. This is critical for business networks with a substantial base, as replacing current infrastructure can be difficult. One way to address this specific challenge is creating a hybrid environment that supports the deployment of virtual networking capabilities in areas with high business value or dire need of upgrades.
See More: What Is Network Behavior Anomaly Detection? Definition, Importance, and Best Practices for 2022
Takeaway
Network function virtualization (NFV) allows for significant scalability and customization by minimizing dependencies on physical network infrastructure. Supporting dynamic resource allocation enables this solution to meet varying customer demands. NFV has the potential to help enterprises increase revenue inflow without a proportional increase in hardware investments.
Did this article help you understand the usefulness of network function virtualization? Share your thoughts with us on Facebook, Twitter, or LinkedIn!
MORE ON NETWORKING
- What Is FPGA (Field-Programmable Gate Array)? Meaning, Working, and Uses
- What Is WPA (Wi-Fi Protected Access)? Features, Versions, and Importance
- What Is The OSI Model? Definition, Layers, and Importance
- What Is a Subnet Mask? Definition, Working, and Benefits
- What is an Intranet? Meaning, Features, and Best Practices
