As the Small Cell industry prepares for the industry's annual get together in London at Small Cell World Summit, it seems that the scope and complexity of what the term "Small Cell" means is being wrenched and stretched almost beyond recognition.
We consider if this is this a good or a bad thing, and what the industry could do about it?
Small Cells have evolved from the early residential femtocell products, when the Femto Forum came up with the following definition:
- Low power wireless access point
- Operating in licensed spectrum
- Using mobile technology
- Generating coverage and capacity
- Over Internet-grade backhaul
- At low prices
- Fully managed by licensed operators
- Self-organising and self-managing
The scope has gradually evolved to include a variety of different technologies
- Wireless access point: Rather than just a small version of a traditional cellular basestation, the scope now includes repeaters, which improve coverage inside buildings, especially in more remote areas where wireline broadband is unavailable. Products such as Nextivity's Cel-Fi are included in a Small Cell Forum document entitled "Alternative Enterprise Small Cell Solutions"
- Higher RF power. Designing products with more than 1W RF power is a different and difficult task. Some vendors specify their Small Cells with up to 5W and I've even seen as much as 20W (the largest that meets macrocell regulatory constraints).
- Unlicensed spectrum using Wi-Fi. I don't see Standalone Wi-Fi access points being included within the scope of Small Cells, but Carrier grade Wi-Fi is often specified as an integral feature of 3G and LTE Small Cells. Various industry proposals for tighter integration include LTE-U (using LTE in Unlicensed spectrum) and seamless roaming to/from Wi-Fi using HotSpot 2.0
- Distributed Antenna Systems and Cloud RAN, where the heavy duty broadband processing is done elsewhere, either at a common central controller within the same large building or remotely at a nearby macrocell or centralised location. Today's products require dedicated cabling for backhaul, anything from dark fibre to dedicated CAT6 cabling, rather than "Internet Grade Backhaul".
- Low prices: Arguably achieved for product cost by huge investments in silicon (Systems on a Chip) and software. The TCO (total cost of ownership) looks good for many indoor use cases, with Urban Small Cell deployment needing to encompass a wider set of costs (site rental, backhaul, deployment).
Aspects I don't believe haven't changed are the need to generate better coverage and more capacity, be fully managed by the cellular operator and use greater levels of automation such as SON (Self Organising Network). If you were to include the addition of roaming Wi-Fi using HotSpot 2.0, where non-mobile technology from roaming partners augments traditional cellular services, then it could be argued that even these bastions of the scope are no longer valid.
Recent analyst commentary has been to advise that operators need to adopt a cocktail of different technical solutions. This is no longer presented as a binary battle between macrocells and small cells, but a more complex pot-pourri where every type of product can find a home.
Analysts concur that after a few years of heavy investments for LTE rollout, especially in North America and China, we may now expect that to throttle back, switching investment into other areas. Short term focus on LTE alongside Wi-Fi- offload will give way to longer term focus on Small Cells in all their forms.
So how does one make sense of this and navigate a way forward?
Firstly, there are some tenets which aren't challenged.
Provision of better coverage and increased capacity will need many more antenna than ever before. These can range from those directly connected to Small Cells to those used by DAS systems. Some might even consider MIMO antenna arrays used with larger cell towers in what are termed Super-Cell architectures.
There is also widespread recognition that Self-configuring and Self-Managing capabilities will be needed to make any HetNet efficient. This includes powerful SON software, which will continue to evolve from the mix of vendor specific and 3rd party specialist components.
There are also a potentially additional requirements entering the fray
Demand for multi-operator capability. Greater acceptance of site sharing is extending down to Small Cell infrastructure. Operators are asking how they can best share the cost of a Small Cell deployment, whether indoor or outdoor. This is less evident for smaller installations, such as in the home or smaller offices. There are also arguments for single operator schemes where local Small Cell controllers can deliver additional services to differentiate.
Closer integration with Wi-Fi. As Wi-Fi continues to evolve, bringing seamless roaming through HotSpot 2.0 and the promise of Gigabit download rates using Wi-Gig, Small Cells have incorporated this capability almost as a standard feature. It's no longer enough to simply include the radio interface – there needs to be a complete end-to-end solution which actively manages use of Wi-Fi to achieve adequate QoS from all available resources rather than the "blind handoff" we often find today.
Scalable deployment and documentation processes. Quite apart from any automated provisioning and configuration systems, operators need to use project management systems which can oversee deployment at vastly increased rates compared to today. These will also require accurate and verified details of the exact location of each Small Cell, handling moves and changes, faults, repairs and upgrades. Many aspects may be outsourced to more than one 3rd party, automation of the contractual processes will have an important role to play.
Some definitive choices
Two key technical factors that will determine the best way forward are backhaul and spectral efficiency.
Small Cells initially targeted the use of "Internet Grade Backhaul", deliberately locating the complex baseband processing within the Small Cell. That's because the cost and availability of high speed/high quality backhaul becomes more difficult to find, provision and maintain as you increase the number of cellsites. While it's true that there is greater availability of fibre and higher performance broadband, this is far from the case everywhere and places an extra constraint in certain locations/use cases.
Spectral efficiency using sophisticated LTE-Advanced techniques. The macrocell industry has worked hard to squeeze the most out of existing assets, with LTE-Advanced using complex techniques to achieve high data density and improved cell edge performance. However these need tight synchronisation using phase timing that can be difficult to engineer. There is a tradeoff between simply installing a few extra Small Cells and precision engineering the system to achieve that with less equipment.
When considering which of the technical options to choose from, the key commercial factors will include:
- Time: Change takes time. Cisco VNI and similar traffic forecasts highlight the enormous capacity growth required in the years ahead. Scaling up and adopting new processes and procedures may take months or even years, and that's before the large volumes of deployment are in place. Some might choose a route involving outsourcing to companies that can move more rapidly, or adopt short term policies that use roaming/network sharing at higher initial cost. Operators who are late and can't offer as good coverage or capacity would be at a competitive disadvantage, resulting in revenue/market share loss.
- Cost: This is never as simple to determine as it appears. It must include the full TCO, consider longer term volume discounts and make comparisons across a variety of use cases. For some technology choices, early adopters might see higher initial costs but would learn most and be best positioned to take full advantage of new technology as it grows to volume. For some applications, such as sports stadia, high costs may be acceptable in order to preserve the brand image.
- Revenue: Many operators have pointed out that they see little additional revenue arising from deploying extra equipment. The business case does need to balance cost with revenue, including the impact of increasing/declining revenues through changing market share. Inevitably more capacity will need to be provided at ever lower cost per bit.
- Risk: Some will prefer to accept a lower risk by using solutions from existing suppliers and slowly evolving existing processes. While this might be lower risk, it may not deliver on the other factors.