If you do not change browser settings, you agree to it. Learn more
The EU cookie law (e-Privacy Directive)
Please visit ICO website for more information. http://ico.org.uk/
|This email contains graphics, so if you don't see them, view it in your browser.|
Connect-World, the information and communication technology (ICT) decision makers' magazine. We are the decision makers' forum for ICT driven development.
|Connect-World's eLetter April 2014||8th April 2014|
SDN - The Dream and the Reality
SDN seems to inspire the whole industry. It is a whirlwind to brush off legacy notions, resulting in a new landscape. Some carriers are projecting on to it unfathomable capabilities of freeing themselves off the telecom equipment manufacturers, with their release plans and high specialist equipment costs. Software defined, not hardware-defined networks would free the world from purpose build network equipment, where the new network will run on commoditized, low-cost white-boxes (unbranded) that have standard throw-away components.
Operators want to be more like the internet ‘moguls’, with ecosystem of developers inventing applications. Services will no longer be designed by marketing departments and R&D engineering teams, but - with the help of orchestration and service modularization that are exposed by APIs, numerous small innovative developers, even the customers themselves, will design their own services. Where services are programmable, not implementable, ‘fast failure’ can be tolerated without incurring heavy losses, and spiky success can be accommodated, with easy expansion of network capacity.
A panacea? Certainly. However, there is a grain of reality there. SDN seems to promise something for everyone in the industry. It is riding on the crest of a wave of ever-increasing data traffic volumes, which is not supported by rising prices to compensate for the increased investment in capacity. Therefore, more efficient utilization of resources and cheaper bandwidth are a lifeline for many carriers.
However, white-boxes are not the main benefit. Easier network management and fast provisioning, where the control is separated from the data plane, is far more effective in reducing costs - lower OPEX rather than CAPEX. SDN is promising to reduce network complexity and integrate legacy as well as new networks under the same SDN controllers, which create logical ‘tunnels’ that are overlaid over the physical infrastructure, wherever there are free resources available for the service. This is not just fast provisioning benefit, but also further automation, where networks can ‘learn’ and adjust without having expert engineers involved. Self-sparing of standard components reduces maintenance costs and engineers’ effort further.
The good news continues. With such easy provisioning, network resilience is improved since capacity is boosted automatically when unpredictable surge of demand is detected. Resources can also be pre-allocated with much greater precision - per customer, per time, location, and service type. Knowledge of the entire network, which was fragmented before, will be available from the SDN controller to the application layer, so that applications can request resources, as and when they are needed - under suitable SLAs. Such ‘Precision SLAs’ empower network providers to provide levels of service at differentiated prices. At long last, there is a way to combats OTT free ride, and justify the increased investment in capacity.
Follow us on
Latest ICT Event
• Africa and the Middle East
• Latin America
• North America
Now, fast-forward to the best news yet. With SDN overlay of control over disparate parts of the network, true visibility emerges that will enhance both network and applications analytics. With SDN-based DPI, consumption requirements can be compiled with great granularity, leading to fine-tuning of service packaging and pricing. Most importantly, this facilitates truly targeted advertising, combining knowledge of subscribers’ demographics with detailed behaviour patterns. Communication policy control, which was only possible on carrier networks, now will be available to corporates that are struggling with Big Data, proliferation of video services and bring-your-own-device.
With so much good news, is it a wonder that the whole industry is galvanized over SDN? No one can argue against such a promising proposition, but it is clear that it will take time, a long time to arrive. First the standards must be agreed, so the dream of multi-vendor networks can be accomplished. This is now underway with the OpenDayLight, OpenFlow and OpenStack initiatives. Integration with legacy will not come cheap or easy. Hence, true visibility and true sharing of capacity will be slow to be realized. The big vendors are by no means ousted yet - they will compete on the best SDN/NFV, best automation, best new management systems. Developers will not flock to SDN networks, unless easy APIs (really easy!) are offered, and there is a fair business model that rewards them. There is much to sort out and many pitfalls, but the gut-feel is that this one is not a passing craze- this one will fly!
|What can Connect-World offer you?
|Copyright © 2014 Connect-World. All rights reserved.|
|If you wish to cancel your subscription to this newsletter,|