Network Functions Virtualization (NFV) explicitly targets the two biggest problems facing network operators: bringing costs in line with revenue growth expectations and improving service velocity. NFV's premise is that industry standard IT virtualization technology (servers, switches, and storage) located in data centers, network nodes, or end-users' premises can be used to reduce the cost and increase the speed of service delivery for fixed and mobile networking functions.
NFV and SDN are related but they address somewhat different business and technical issues. NFV was launched by network operators to reduce network cost and increase service velocity of public networks. SDN was conceived by data center operators to address data center network complexity and cost issues. NFV addresses fixed and mobile networking functions such as fixed access, home, and IP mobile core networking functions. SDN originally addressed physical and virtual data center switching but now addresses low layer network transport, switching, and routing services in all networks.
Recently the Network Functions Industry Specification Group defined NFV use cases for many network functions. Home network, fixed access and content distribution network (CDN) virtualization have the potential to fundamentally change fixed network economics. (I have included numerical results from some of my recent business case studies in the discussion of each of these use cases below.)
Home network virtualization replaces the set top box, including storage for personal video recorder and residential gateway, with a simple low-cost on-premise device driven by cloud-based functionality. The set top box today provides media streaming services, video on demand, cable guides, and rudimentary web browsing functions. A set top box is required for each television. The residential gateway provides connectivity, security, and limited subscriber management. Set top boxes with this level of functionality cost in excess of $100 and typically require installation costing an additional $100. The residential gateway costs $50 to $100.
In the U.S. alone there are 115 million cable subscribers and about 250 million televisions. The cost to move to a new generation of set top boxes, therefore, is about $20 billion to $40 billion. Such initiatives must be spaced out over many years. The high cost and logistic challenges of upgrading home networks, consequently, is causing fixed network operators to lose ground relative to OTT and mobile network providers. NFV breaks through these cost and logistics barriers by employing self-installed hardware priced in the $25 to $40 range driven by highly scalable cloud-based software. Software releases are deployed in virtualized data centers rather than rolled out to many different models and vintages of set top boxes. This cuts service innovation down to weeks rather than years. Some of this capability already exists. In a recent study I found total cost of ownership over five years can be reduced by 83 percent through use of a cloud-based cable guide solution versus replacing existing set top boxes.
Conceptually, the virtual fixed access use case is not that different than the virtual home network use case, nor is its business benefit. Millions of costly devices mounted on or near homes and businesses are replaced with hardware ranging in cost from $50 to a few hundred dollars while the high-value and high-cost functionality is moved to highly scalable and low-cost virtualized data centers. Capex is reduced by at least 5:1 while opex is reduced by at least 7:1 according to my recent studies. Also, like home network virtualization, new services and innovations can be deployed systemwide within weeks versus the multi-year rollouts required with existing technology. In addition, niche service offerings are made feasible by dramatically lowering the time and cost to create and deploy a service offering. Many small opportunities can be launched within the time and budget formerly required to launch one major service. For example, content specific bandwidth and QoS controls allow a content provider to offer a specific video program (asset) with a faster download time and better picture quality at a premium price. In turn, the content provider pays the network operator a delivery fee on a per view basis.
The virtual content distribution network use case separates a cost-reduced CDN system deployed at many points within network operators' metro networks from cloud-based CDN functionality. This creates leverage for the network operator by enabling a shared CDN service where many different content providers operate autonomously over the network operator's CDN hardware. Pushing CDN out toward the end-user frees up a great deal of network capacity upstream from the CDN device because unicast video streaming dominates metro network bandwidth requirements. A probability model of likely video streaming requests found that aggregation network and Internet peering costs were reduced by 45 percent to 55 percent by pushing CDN further out into the local network.
Game-changing benefits can be realized by targeting NFV use cases near the tens of millions of fixed network end-points. This is where the money is. Also, moving network functionality into virtualized data centers lowers the many technical and distribution channel barriers that are retarding fixed network innovation. The Network Functions Industry Specification Group's advocacy of standards based solution also will help to open up the network to a broad ecosystem of innovators.