Over the next few years, we can expect operators in developing countries to adopt Network Function Virtualization (NFV) as a key concept within their network infrastructure.
The transition to IP Multimedia Subsytem (IMS) architectures is now underway as operators look to offer more than vanilla voice and SMS and embrace a new generation of IP data services. This provides an excellent opportunity for operators to start virtualising their networks because they need to buy new network functions for their IP data services. Virtual Network Functions (VNFs) are the obvious choice – fast to deploy and eliminating expensive, un-used capacity, whilst offering the flexibility to continually adapt to market changes.
NFV allows network elements and ultimately the entire network to be virtualised and hosted in public or private data centres. Traditionally, network functions have been hosted on purpose-built hardware and software platforms and their functionality was hard-baked into the system so that changes could only be made by the equipment vendor. NFV can put the operator back in control, with network functions hosted in a cloud-based environment which is inherently flexible and cheaper to deploy and develop.
As operators integrate VNFs into their networks, one of the greatest benefits they gain is the ability to deliver the required capacity of their network functions as it is needed. Traditionally, operators have had to dimension and purchase sufficient capacity to service the peak load for that function. The peak load, for example during religious holidays and New Year, would often be many times as much as the typical daily load. These factors mean the system has un-used, idle capacity most of the time and the functions it provides are rigid and unchanging.
Using NFV, rather than dimensioning the function for the peak hours of use, operators can add more NFV functions or reduce the computing resources as required. They can profile the capacity they typically need during normal operation and only use the extra capacity for peak periods of activity. Traffic conditions can be monitored so that traffic levels that exceed the anticipated traffic profile can be recognised and additional capacity brought online in good time.
The approach will also assist when extraordinary events such as a major incident emergency occurs. Whereas a traditional fixed capacity system could not cope (and could very easily fail to operate for any of the prospective users), a VNF can bring additional resources on line relatively quickly to service the extraordinary need.
VNF capacity profiling need not be limited to a carrier in a single country; there are tremendous economies of scale to be had from deploying a centralised, virtualised network once for use across an operator group operating in several countries. If the countries operate in different time-zones, then so much the better, as their respective peak periods will probably occur at different times. For operator groups which use their own data centre, this would be particularly beneficial, as it would reduce the un-used capacity.
Careful use of VNF computing resources enables operators to reduce Capex and Opex and run their networks more efficiently. This is key for the success of operators in developing countries, searching for profitability in regions where revenues and margins are tight.
An open door to innovate
Another benefit of NFV is the DevOps-style business model that it enables. In many developing markets, smartphone ownership rates are higher than computer ownership due to a new range of smartphones costing under £30. As such, local mobile operators often offer services that are central to the everyday life of their customers – such as mobile banking, health and medicine services, and commerce apps. Such operators are therefore supplying the entire computing and communication infrastructure and there is an end user expectation of continuous enhancement to the applications they use. By using NFV to deliver these services, carriers in markets such as Africa can move to a “web-style” DevOps model to meet these customer needs.
NFV helps with service testing too. Most network operators have a separate test network which is used to test changes prior to rolling them out to their customers. Using NFV, operators can create multiple cloud-based replica test networks as and when they are needed and avoid the inevitable bottleneck of testing environments. By using NFV operators can develop, test and launch new services much more easily, reacting to the ongoing need of their customers, whilst also targeting new customer segments.
Furthermore, service development and innovation can also be opened-up to third parties. External developers can be engaged to create applications and solutions for specific regional markets in accordance with local regulation far more efficiently than remote, global hardware vendors can. Tailored services can be developed and rolled out to multiple countries or regions from a single network location, and scaled as demand increases. The result is that consumers and businesses across the developing world could see entire ranges of services created specifically for their needs.
Thanks to smartphones and the move to all-IP, telecommunications is experiencing a period of rapid and continuous change. By taking advantage of NFV, carriers across emerging economies can scale their networks and services according to actual customer demand, rather than relying on old business models, projections of service take-up and up-front purchase of capacity. Service differentiation is a realistic possibility, and by offering and innovating through NFV, customers can be served much better than before; boosting local economies and continuing to create a better connected developing world.
Chris Haddock is the head of marketing at OpenCloud.