Implicitly Unstable Networks - SK Telecom Dares to "Drive by Wire"

Posted by Scott Sumner on Wednesday, July 15, 2015 with No comments
Are we Nearing Auto-Pilot for Networks?
The SK Telecom announcement we made is underpinned by a larger industry trend, which is the migration to truly intelligent, self-organizing, optimizing and machine-controlled networks.



This trend parallels a number of other changes in the way engineers see risk and stability as no longer implicitly required by an underlying, physical infrastructure. We are accepting that control requires automation. Take flight, for example: today's fly-by-wire planes are inherently unstable, and in fact cannot be manually controlled if the flight control computer is impaired. The same is true for self-driving cars: they will crash without visibility and automation. This is a big step from backup cameras--a 'nice to have' feature that will soon have real safety implications without the driver's involvement.

As mobile network traffic, technology and performance requirements continue to escalate--driving the need to move towards faster 5G networks--mobile network operators also have to shift to software defined networks (SDN). This fully automated, self-optimizing infrastructure is required to maintain customer experience and network performance visibility in the major transition to virtualized networks (NFV), while new applications like 4K video steaming and IoT start to fill their networks.

Until now, SDN has been primarily used in data centers, and only implemented in a handful of communications service provider networks. However, the desire to have cloud-like agility, network programmability and agility, combined with the cost savings network virtualization offers, have all major operators--AT&T (Domain 2.0 program), Deutsche Telecom, Vodaphone, CenturyLink, Telefonica, Softbank and many others--in a race to realize these benefits over the coming few years.

SK Telecom's big data analytics-driven, fully programmable SDN network is an exceptional achievement, realizing a large-scale, fully automated network on the level of Google, Amazon, and other major cloud players, in a network delivering carrier-grade services to over 28 million subscribers. The integration of performance monitoring as a nervous system that drives the network to adapt and optimize itself is a highly significant addition to SDN strategies, and is key to the industry as it validates the strategies all other operators are embarking on--mostly on faith in the return of SDN and NFV as new, nascent technologies that have never been proven in a large scale deployment.

Seeing the success SK Telecom has realized in implementing "the networks of the future" now, not only validates the strategic plans of most top-tier network operators worldwide, but gives their teams, their subscribers, and their investors more than simply hope that they will soon reap exceptional benefits in revenue growth and operational efficiency.

Accedian solutions are intimately integrated into the SK Telecom network architecture and automation platform, and the Accedian executive team can provide unique insight into their big data maturity model, and the benefits this has driven across the largest Korean operator's network, and what will be a key differentiator for all other providers in the coming years.