In the long term, the industry generally seems to be agreed that Small Cells are the way to go – almost every operator has these on their roadmap – but they've not taken off in volume yet. Operators continue to shout loudly for more spectrum to be made available, but this isn't always the easy option to deploy quickly and effectively that it might seem.
Fundamentally, there are three ways to increase network capacity:
- Add extra spectrum, often auctioned at high prices by governments but (free) Wi-Fi spectrum is also seen as an attractive option by some.
- Squeeze more user data into the same spectrum more efficiently, by refarming existing 2G and 3G bands to use 3G and 4G
- Reuse the spectrum many times over, splitting sectors and adding new cells
A number of major initiatives common for most mobile network operators
A series of large strategic investments we've seen from several operators typically include the following major milestones:
- Single RAN basestation refresh
- Refarming of 2G spectrum (e.g. using 3G at 900MHz)
- LTE launch, typically using new spectrum
- Higher speed LTE, either through wider spectrum (20MHz bands) or Carrier Aggregation (e.g. 10MHz + 10MHz) or both
- Adding further spectrum for LTE (e.g. 2600MHz, 700/800MHz)
A common theme throughout all these upgrades is a growing focus on perceived network performance by the end-user. Many operators recognised a few years ago that the high data rates achievable through 3G weren't translating into the rapid and responsive end-user experience on smartphones.
Dedicated engineering teams worked through the end-to-end service provided, measuring this for the most popular applications (Facebook, YouTube etc.) and optimising performance. Low latency for 4G has been particularly important and dramatically improved perceived application speeds when properly engineered.
Other large projects which differentiate
Site sharing deals, sometimes full radio network sharing which reduce the total number of cellsites but make each site more complex. These are all driven by the need to drive down costs, avoiding duplication and unnecessary expense.
VoLTE (Voice over LTE) launches are delayed in many markets for various technical reasons. Some operators, such as Sprint, have decided instead to introduce HD Voice over 3G and defer Voice over LTE until it is more mature. Others may choose never to adopt it.
Wi-Fi offload, where different operators have taken radically opposing views. All now ensure that every smartphone sold has Wi-Fi capability, which some then ignore while others seamlessly integrate it into their own and/or partner Wi-Fi networks.
Adoption of other technical solutions varies. Some seek out advanced technologies to increase spectral efficiency, such as higher order MIMO (requiring additional antenna at each site) or Cloud RAN (requiring dark fibre from each site to centralised baseband datacentres).
Adding spectrum isn't always straightforward
Macrocell sites adding new spectrum bands typically need additional antennas to handle those frequencies. Sometimes these can be shared with other operators, while other cell towers are positively bristling with new ones. Issues include physical space to accommodate them and whether the towers have enough strength to accommodate not just the additional weight but stress in high wind conditions. A new physical tower may be required. In some situations, planning permission for a larger installation may not be feasible or would at least be costly and take time. More sophisticated multi-band antennas may be a workaround but also attract further expense.
A secondary factor is that different frequencies would have a different range and footprint. Sometimes that's a good thing – new 4G service at 700MHz sharing towers previously used for 3G at 1900MHz should reach around many more corners and deeper inside buildings. But 2600MHz would have less penetration than1800MHz and this could be an issue when considering Carrier Aggregation services that use both together.
Where antenna are shared, the RF combiners, cabling and other ancillary components may also need to be changed. These can be quite costly when compared to a standalone fully integrated small cell.
A third consideration when adding new spectrum is the compatibility of smartphone devices to support it. While the very latest products may do so, these typically only comprise a very small percentage of the total installed base and total traffic used. A relevant case study for that was given by KPN where they considered whether it was more cost effective for them to identify and target early free handset upgrades to specific customers rather than deploy extra spectrum at specific sites. There can be a period of 1-2 years before customers transition to devices that can make best use of the spectrum available for deployment.
Sector splitting or small cells
An alternative method of doubling capacity is to split sectors that are highly loaded, for example a typical 3 sector site can be reconfigured into a 4, 5 or 6 sector. Again, more antenna are required (sometimes causing planning or landlord issues) and onsite equipment, and this can't be repeated once the new capacity is reached.
Adding outdoor urban small cells (metrocells) achieves a similar result, but adds significantly more capacity. One view I've heard is that it may not make sense to add a single urban small cell to an overloaded sector – sector splitting would achieve the same result – and the scenario of a cluster of small cells deployed in a busy urban or suburban area is much more likely. I don't believe this is a cut and dried strategy – there will be cases where service performance is better achieved by a well positioned small cell.
The key factor here is how well prepared the operator is to handle deployment of small cells . Have they made the upfront investment in time and effort to setup the processes to deploy, manage and integrate small cells into their network cost effectively?
Indoor resources offloaded may achieve disproportionately higher benefits
It's easy to look simply at figures of total traffic throughput to measure the utility of an individual cellsite – how many call minutes and/or GBytes of data are consumed. However, the resource required to deliver a Gbyte of data from outdoors to inside a modern building can be much greater than when serving users outside – sometimes 10x or more – because the modulation rates and packet loss are dramatically different.
To me, this makes the installation of in-building small cells even more urgent than outdoor. Other factors include the lower costs – often the building owner will provide the site and power free of charge, and may even contribute towards the cost of the small cells and/or installation themselves. There is also the added benefit of much improved in-building performance which can attract and retain long term business and consumer customers. Offloaded traffic releases those expensive resources in the outdoor macrocells and defers the need for further capacity investment.
Planners will have to make some commercial calculations to make those trade-offs both at strategic long term business direction as well as at the tactical levels. The sales channels and enablers for in-building small cells are quite different to those for outdoor scenarios, requiring some up-front investment in processes and procedures to support them.
Find ways to offload the most resource hungry data, including Wi-Fi or in-building small cells, and establish strategic initiatives to make that happen.
Compare the costs of adding a sector/antenna/baseband with deploying a cluster of small cells. Is this a short term fix or a longer term solution? Assess the maximum limits and deployment timescales for both. Establish groundrules when making tactical decisions.
Clusters of metrocells deployed for capacity may be equally relevant in suburban areas as for urban. The attraction of LTE for some means that users may prefer using it over Wi-Fi, even where it costs more. Assuming that network operators have priced their service properly, the additional revenue should be welcomed. Establish procedures to identify those hotspot areas and introduce plans to deploy small cell clusters in line with trending data growth patterns.
A closing thought – just as some operators are choosing to delay/defer VoLTE rollout, we've also seen examples of declining to bid for new spectrum on offer. Maybe you can have too much of a good thing after all.