SON (Self Organising Networks) is a very active industry topic at the moment. It's seen as an essential enabler for HetNets, where the complexity of co-ordinating between multi-layer, multi-vendor, multi-technology radio networks needs sophisticated automated configuration to work effectively.
One large US operator has publicly stated they plan to grow from 100,000 active cellsite sectors today to more than 500,000 over the next 5 years (includes Small Cells but excludes residential femtocells). This rapid rate of increasing complexity in network management isn't uncommon and demands sophisticated automated configuration and optimisation systems.
Quick recap on C-SON vs D-SON
SON comprises separate functions distributed throughout the basestations (whether Small Cells or Macrocells) called D-SON (Distributed SON). This includes real-time reaction to current traffic demand and environmental conditions and is embedded by each Small Cell and/or macrocell vendor within their equipment.
C-SON (Centralised SON) performs more of an orchestration task, supervising and gently refining the end-to-end system performance. The major RAN vendors also provide C-SON functionality, which is likely to work best with their own systems.
An Infonetics survey in 2013 found that over 80% of operators worldwide have deployed SON commercially, with 3G being the primary use case.
Independent vs RAN Vendor SON
There remain several independent SON vendors without their own RAN equipment who may be seen to have an advantage because they have to work across all vendors. On the other hand, RAN vendors may take advantage of proprietary features and inside knowledge of their own systems to optimise their own end-to-end systems. Operators will need to consider whether the perceived reduced risks of working with a single RAN vendor outweigh the potential cost and performance benefits of a multi-vendor network.
We've seen a number of acquisitions of SON vendors in the past year, such as Cisco purchasing Intucell, TEOCO buying Aircom and Amdocs snapping up Celcite and Actix. The remaining independent SON vendors are becoming more visible and expanding their market scope.
Start with the low hanging fruit
The initial C-SON deployments were targeted at cost reduction. It's still quoted as the primary driver for the SON business case. It's easily seen that if you could squeeze just a few more percent performance out of an existing radio network that had cost (say) $10 Billion to build, avoiding or delaying additional CAPEX to build extra cellsites, then the cost of the SON system quickly pays for itself. Add OPEX cost savings by not needing additional skilled staff to handle the growing complexity of the network adds a further commercial benefit.
Popular use cases for C-SON include CCO (Coverage and Capacity Optimisation) which concentrate on tuning the RF power transmit levels and configuring the neighbour lists in each cellsite to optimise performance. Data traffic may be shifted between available frequencies and sites to dynamically balance the traffic load to available capacity.
Today's focus to optimise the existing macrocell layer is extending to prepare the way for growing numbers of Small Cell deployments. It's essential that operators have both confidence and experience in their SON systems before integrating large numbers of additional cells onto the network – especially those outdoor which will need even tighter co-ordination.
As the deployment rate of Small Cells grows, SON becomes a key factor to facilitate rapid rollout – without it, the manual configuration processes for new cellsites simply wouldn't scale. Evidence of this can be found in Korea, where operators such as SKT used it to achieve deployment rates of 100 Small Cells per day or more.
SON as a revenue generator
I spoke with Ofir Zemer, CEO of Cellwize, one of the larger independent SON vendors. He reports the view that operators don't see SON purely as a cost saving measure for OPEX or CAPEX reduction - it's also about generating more revenue. Cellwize want it to be more visible that SON translates into better ARPU (Average Revenue per User) and higher lifetime value, believing that this will become more pronounced as the industry matures. He argues that improved network performance and effectiveness increases both usage and customer loyalty.
They confirm that splitting SON functionality between D-SON and C-SON is appropriate. It makes sense for some SON actions to be taken near the element to ensure a rapid response while others are better done centrally on the basis of a larger data set.
Ultimately, there must be a "Master C-SON" for any geographic area that determines how the system is configured, but it is possible for larger networks to use different SON systems in different parts of the country or in different cities.
Growing activity in the form of many RFPs (Requests for Proposal) for SON solutions today is indicative that operators are preparing their network management systems for added complexity and scale.
This will be essential aspect to prepare for mass deployment of outdoor Small Cells that are tightly integrated with the rest of the radio network. This speeds deployment through self-configuration and optimises end-to-end performance of the network.
Benefits are not limited to cost saving or Small Cell deployment speed/capacity but can also be viewed as revenue generating. Customers value the higher quality of network performance which is becoming more visible as data traffic loads continue to stretch and fill up available capacity.
ThinkSmallCell will be speaking on "SON: The Future of Network Management" at the Network Management Show in London, 2 July 2014