Towards the self-driving telco – SCF’s roadmap for SON and orchestration

Densification will mean very large numbers of cells being deployed – 5G is likely to have at least five times more cells than LTE, and in areas of high usage, the figure could be 20 times or higher. This raises many challenges in how such large numbers of elements can be constantly optimized to deliver the best performance for any given application.

Automation, with technologies like SON (self-optimizing networks) will be essential, and as operators move to SDN (software-defined networking) environments, small cells will be able to be orchestrated flexibly by network-wide controllers.

These two key technologies, SON and orchestration, will be vital to avoid chaos in the dense network and ensure that large numbers of cells can work together to deliver excellent quality of service (QoS). 

This is why these two topics are at the heart of SCF’s work program for the year ahead. The TECH & 5G working group is embarking on a vital project entitled ‘Roadmapping SON and orchestration’.

 This comes at exactly the right time for the industry. We know that MNOs are already making difficult decisions about 5G, densification and SDN, and how those impact on one another. They need clear guidelines and rich information to help make their plans and be confident about their success.

Common technical frameworks and open interfaces will help avoid reinventing the wheel and the SCF work will build on ground-breaking activities which have already taken place in these areas, including the SON API (application programming interface), which allows multiple SON systems to work together; and the HetNet Architecture Framework, which has a heavy emphasis on automation and the road to SDN.

 As well as building on those foundations, new frameworks and blueprints will also draw on best practice from early adopters of SON and orchestration in the small cell environment, whose sharing of their cutting edge experiences is vital to help other operators make informed decisions.

 All this will become critical soon, because operators are embarking on large-scale densification. With so many elements in play, it will be essential to automate most of the processes of planning and optimizing these dense networks. If every change in a cell’s condition, or users’ behavior, has to be manually addressed, it will be impossible to deliver the highest quality of service.

 Automation can be enhanced when small cell networks are virtualized, but that involves further important architectural decisions.  Virtualization will be an increasingly common way to handle huge numbers of access points efficiently. Clusters of cells in a high usage area can share a centralized controller, which provides most of the baseband processing for the whole cluster – and sometimes power, backhaul and edge compute resources too.

The baseband functions in this scenario will be virtualized – run as software on the shared controller – which allows network resources to be allocated far more flexibly around the cells, according to traffic patterns.

This move also allows the small cell network to be managed more flexibly and proactively within the wider operator platform. In an SDN environment, all network assets, virtual and physical, can be orchestrated in a common way, to maximize efficiency and user experience. The orchestrators can allocate and re-allocate resources as requirements change, to ensure that every user has the best performance and critical applications are prioritized.  

With such huge numbers and variety of elements – and another sharp leap driven by the IoT when that achieves mass scale – the management and orchestration (MANO) of the HetNet must be almost entirely automated, and must run from end to end to coordinate every one of the resources.

Therefore, the fully automated, ‘self-driving telco network’ is one of the cornerstones of the next generation vision for many operators, since it would keep track of all these moving parts, enabling them to be allocated and provisioned optimally and efficiently for each user and application.  

That is why this will be such an important priority within SCF’s work program for 2018-2019. We will be working with other organizations which are focused on SON and MANO, such as the Open Network Foundation and ONAP (Open Network Automation Protocol). The results of the work will undoubtedly influence standards bodies, open industry frameworks and individual MNO strategies in the years ahead.

To have first access to the ongoing results of this comprehensive work, and to make your own contributions, why not join SCF today?

Caption: SCF’s 10-year roadmap for SON, orchestration and automation. Source: SCF TECH & 5G working group (for members – you can find more TECH &5G Group documents on the members’ site, under file share)