SDN and NFV are now a strategic consideration for operators, but the network and services are still evolving. Facilitators that enable change, SDN and NFV can determine what is to be orchestrated in the network but there is another critical stage that is missing – the intelligence to decide what changes need to be made to the network, or the service placement, and when.
In this blog, Dr. Jay Perrett, chief technology officer, Aria Networks discusses the need to create a ‘closed loop system’ that can automate network operations; necessary if business driven optimisation is to be realised.
Taking the Gartner Hype Curve into account, SDN and NFV are firmly in the phase of serious consideration with operators already making strategic decisions on using the technology. It is therefore important at this stage to ensure that all aspects of the network, and associated service evolution, are carefully considered to make an investment that also provides a robust working solution well into the future.
So, if SDN provides the separation of data and control and NFV provides the separation of hardware and function, what is left is the frequently overlooked requirement that is the need to create a closed loop system that will ultimately automate network operation. Put simply, Artificial Intelligence is the separation of people from engineering. Aria Networks believes that by putting intelligence in the orchestration level the network can then decide WHAT the changes needed should be, WHEN they should take place and WHERE. Combined, this will deliver faster network decisions and improved operations resulting in improved time-to-market for service delivery.
When networks finally have superior levels of intelligence embedded in them a significant shift in network management can occur. The role of the human network designer, or operator, can be automated and programmed to deliver not only in a dynamic nature but also at the high speeds demanded. It is important to note here that this does not mean that humans are removed from the equation. A person is required after all to clearly define the business objective to which a network and its exposed services must operate. In some cases (such as a service level agreement with a customer), this might be primarily based on timely service delivery. In other cases (such as non-essential backups) this might first consider cost. The most intelligent networks will be able to consider multiple criteria, even with seemingly conflicting objectives, and deliver an optimal network orchestration.
A future in which super intelligent networks will self-optimise is not far away and the benefits to the service provider here are clear; better optimisation of resources, more predictable current and future investment and the ability to confidently provide service level agreements to customers. In addition, operators will be able to effectively predict network behaviour ahead of a significant product launch, or before introducing new promotions, and subsequently strengthen any areas of potential weakness ahead of the rush of users all vying for access at the same time – and all without any direct human intervention.
Whenever Aria Networks talks to a customer, or an operator, it is ultimately clear that the CEO or CFO wants more control over margin and network management so just how much longer will it be until the benefits of using Artificial Intelligence to optimise networks is universally applied?
The author of this blog is Dr. Jay Perrett, chief technology officer, Aria Networks.
Comment on this article below or via Twitter: @ VanillaPlusMag OR @jcvplus
