In IT, the value proposition in moving to the cloud lies in the sharing of resources and creation of a “pay-as-you-go” business model. This is similar to the pricing used by utilities, which charge customers based simply on the amount of electricity, water, or gas they use.
Operators are rethinking their business models to be more in line with this approach. But their current networks are generally a heterogeneous collection of equipment and architectures from different vendors, which makes it much harder for them to handle surges in customer use, or to shift resources from one area to another.
In response, network operators are adopting new cloud and virtualization technologies in order to capitalize on software-defined networks (SDNs) and network function virtualization (NFV). According to an IBM survey, adoption of cloud/virtualization in the industry has grown from 11% to 40% over the past three years; the number of companies has grown from 28% to 40%.
Moving to SDN/NFV offers network operators far greater flexibility in deploying various functions with the added bonus of working in an open, standardized environment. A truly virtualized, cloud-based network architecture gives both hardware providers and mobile network operators the ability to efficiently operate, maintain, and upgrade network resources while speeding up functional and service deployments. It lowers CapEx by reducing the need for specialized equipment and tool chains. And it speeds innovation, delivering new and enhanced services (and revenue streams) while re-using existing equipment.
These new technologies help simplify network security management, such as network access control. In turn, they are making it easier to set appropriate privileges for users and devices accessing the networks, with those privileges following the user or device as they connect from different parts of the network.
Another benefit is the ability to create dynamic links between locations such as among data centers, between the enterprise and data centers, and among different enterprise locations. This infrastructure allows operators to dynamically apply appropriate Quality of Service (QoS) and bandwidth allocations to the links.
Virtualized core systems for service providers include support infrastructure such as vIMS (virtual IP multimedia systems) and vEPC (virtual Evolved Packet Core), which is a framework for providing converged voce and data on a 4G-LTE network, as well as dynamic mobile backhaul, virtual PE and NFV GiLAN infrastructure. Using vEPC and vIMS makes it possible to run EPC software on commercial off-the-shelf equipment instead of proprietary hardware. This also means the mobile core can be run on small, efficient servers in rural areas. Virtualization also helps creates new base business, new product lines, and new services – a major consideration as new devices come to market and the Internet of Things expands online.
The biggest challenge facing operators is figuring out how to adopt these new technologies. Many telecoms try to use internal resources to design these new virtual or cloud architectures. But engineers who think about software-defined networks or virtualized functions the way they think about existing hardware analogs may make design mistakes, create unintended bottlenecks, and introduce weaknesses in network security. Lack of well-designed architectures from the beginning – especially security-centric architectures – lengthen time to adoption and drive up costs.