Even further in the future, there is a chance that, with SDN and NFV, virtual telcos could emerge. These organizations would instantiate services in the physical network of a partner, which they would then market and support in a virtual manner. Operators’ network architectures as they currently exist are incapable of accommodating this use case. Traditional architectures are built in a batch model, linear by nature, passing instructions and data from layer to layer in succession; the architecture of the future is elastic and non-linear, and takes into account how resources are consumed in the environment.
Many analysts, journalists, and market players believe that the telecom industry is indeed on the precipice of a paradigm shift, and that SDN and NFV will transform the industry. (I humbly submit myself as a believer in this revolution.)
The hype around network virtualization is considerable, but Strategy Analytics predicts that “SDN is likely to move even faster than the hype in 2014.” Some operators are already implementing SDN in select domains, such as cloud services or MPLS backbone. Simon Osborne, CTO, Fulfillment at Comptel, told Pipeline that his company is “starting to see requests for model SDN infrastructure in inventory RFPs.”
NFV is moving slightly slower. Michael Howard, co-founder and principal analyst for carrier networks at Infonetics Research, wrote in April that, “it won’t be until 2015 that we’ll see commercial [NFV] deployments kick into motion, still most likely on a limited basis, as operators put one or two use cases to the test under real-world conditions in their live networks.”
How quickly will these trends grow in the coming years? ABI Research predicts that the advent of SDN and NFV will create “a virtualized telecom market of $5 billion to $6 billion by 2018.” Clearly, there is work to be done in the vendor community to accommodate the future.
An orchestration overlay that evolves with the metered introduction of SDN and NFV and supports a multi-vendor, multi-implementation environment is required. In a “living services” environment, service topology and assurance is quite different and must be transformed. Additionally, vendors should consider the amount of data that is generated from SDN controllers around flows that can be captured and monitored, and examine opportunities for leveraging analytics and intelligence around service use.
Rather than looking for the big bang, CSPs would be wise to adopt “platforms and partners that allow them to manage the revolution in an evolutionary way,” says Osborne. The move to SDN and NFV supports the holistic, non-linear view of the management platforms needed by operators in the future, but it also challenges existing architectures, supplier hierarchies and associated market dominances. Currently, IT vendors are pitching virtualization stacks, NEPs are promoting SDN controllers, and there is a pressing need for solutions that connect these disparate virtualization tools into existing OSS and BSS stacks. If CSPs consider these dynamics from day one, instead of making decisions based solely on ephemeral CAPEX and OPEX prognostications, they will be better positioned to compete in the future and their business will run with the same relative ease that my PC could when emulating virtually any gaming platform.