We chose the title - Stretching the Limits - for this UK industry specialist conference to focus on current thinking of how best to squeeze extra capacity from cellular networks. With 5G still some years off, would the consensus be entirely towards Small Cells or would other technologies factor into the mix? Presentations from analyst, vendor and operator speakers combined with an interactive audience Q&A session to deliver an interesting snapshot of current industry thinking.
The mix of more than half a dozen mainstream solutions available will offer different returns over the next 3-5 years, leading to a transition from today's mainstream LTE macrocell deployments towards a greater focus on Small Cells.
There's no single solution
First up was Caroline Gabriel of Rethink Research who outlined the various technology options available today:
- LTE Upgrade
- Wi-Fi Offload
- Public access Small Cells
- Deconstructed RAN (using Remote Radio Heads from existing macrocells)
- New/Refarmed Spectrum
- SuperMacro (using LTE Advanced Features such as CoMP, MIMO etc.)
- Adaptive Networking (using SON)
The table based on market research of current operator thinking showed that the greatest gains this year can be had from LTE deployment, reflecting the heavy investments made to deploy mostly using newly acquired LTE spectrum. Offloading to Wi-Fi and public access Small Cell deployments follow closely behind. Looking ahead to 2018 though, the quick win from LTE drops to 4% and it's clear that the most effective capacity increases will be delivered through public access Small Cells. These have to be a key strategy element for mobile network operators worldwide. They also have to be implemented cost effectively.
A new acronym (to me) was introduced – HetSnets (HetNets with Small Cells). It does reflect that the long term architecture of mobile networks isn't a binary choice between Small Cells and/or Macrocells. We can expect many of the techniques above to be incorporated and work together. Take-up of some of these technologies will be skewed toward specific countries, such as those needing dark fibre to the cellsites.
Caroline also pointed out a few disruptive areas worth keeping an eye on:
- Under utilisation of TD-LTE. Many operators have TD-LTE spectrum but few have announced plans to use it
- Use of higher frequency spectrum, such as Wi-Gig (Wi-Fi at 60GHz)
- LTE-U (Using LTE across unlicenced Wi-Fi frequency bands)
A key takeaway from the presentation was that SON (Self Organising Networks) will be essential to co-ordinate between these different network layers. It was said that SON is the most important feature for Small Cell vendors to focus on today.
Atul Roy, Network Strategy Manager at EE, outlined a clear focus on LTE
EE defined LTE Small Cells as any indoor or outdoor cell that wasn't a macrocell, ranging in RF power levels from 100mW to 10W. These included solutions for indoor coverage, indoor capacity, dense urban outdoor zones and even rural areas.
Resilience plays a major part in their thinking. We all want "Always On" service and that means designing in resilience at various levels – in the RAN, the Core Network and end-to-end architecture. Where outages do occur, the system needs to be self-healing so that it can fall back on neighbouring or wider area cells.
Key aspects that EE would be looking for in any Small Cell solution include:
- Overload and congestion control, so that all available resources are fairly apportioned to end users, giving a reasonable service to all
- Small form factor. "It's not a beauty contest, but form factor matters"
- Cost: Especially deployment and management costs.
- Resilience: As above, more components across the network leads to a higher chance of individual component outage
- Standardisation: The ability to use different vendor solutions seamlessly.
While EE didn't discuss 3G in their presentation, it was clarified that indoor 3G does have a place. EE currently offer a residential Small Cell for those with poor indoor coverage. For in-building/Enterprise locations, RF attenuation reduces concerns about interference with the macrocell layer and allows 3G to be used.
HetNet technology explained
Mike Lawton of Agilent (whose testing division is soon to be rebranded as KeySight Technologies) explained in detail how some of these advanced LTE HetNet techniques work. This included eICIC and ABS where the macrocell stops transmitting and allows time periods for the underlying Small Cells to use the same frequencies.
He reported that much of this technology was being actively tested in the labs today, with early field trials and commercialisation set to follow. It will probably be 2015 before we see this widely deployed, with South Korea and Japan perhaps being amongst the first.
Automating the Network
Two speakers from Amdocs rounded off the presentations. Justin Paul discussed how many operators need to find ways to efficiently scale up their cellsite deployment processes. Amdocs have been working closely with at least one North American operator to project manage the lifecycle of Small Cell rollout. Back office systems previously used to deploy a handful of new cellsites per day now need to be tracking hundreds.
Neil Coleman, also from Amdocs, highlighted many of the SON (Self Organising Network) features that will be critical to simplify and automate Small Cell deployment.
I heard from several speakers and delegates that SON is one of the hottest topics today, with many RFPs in progress. Operators need to get these solutions in place with their existing macrocell networks first, then extend their use to accommodate Small Cells. Confidence in the reliability of SON is growing, but in a risk-averse business it may take some time before manual control is fully relinquished to the software systems.
Shared spectrum vs dedicated Small Cell spectrum
Zahid Ghadialy led a lively and informed debate about whether it was better to allocate dedicated spectrum for the Small Cell layer, or reuse/share it with the macrocell. In most situations, delegates felt that shared spectrum was the most efficient. LTE already has many features built-in for HetNet operation to enable this, such as ABS (Almost Blank Subframes) and ICIC (Inter-Cell Interference Co-ordination).
Over the next few years we can expect LTE Small Cells to be tightly integrated into mobile networks. The emphasis will shift from initial LTE macrocell deployment, spectrum re-farming and Wi-Fi offload over the next 3-5 years. A variety of techniques including Super-Macro, DAS and Wi-Fi will all have their place – some parts of the world (e.g. where dark fibre backhaul is prevalent) may prefer to use more tightly co-ordinated technologies, but everywhere can expect to see many more Small Cells.
The next CW Small Cell SIG event takes place in London on 1st October.
Slides from the presentations can be downloaded freely from the CW website.
David Chambers of ThinkSmallCell is a CW SIG Champion, voluntarily co-organising these events.