DAS and Small Cells have in past years been positioned as strong competitors rather than collaborators. New technical solutions, new opportunities and new thinking are radically reshaping that perception. We consider some of the recent product announcements and regulatory changes which affect that view.
Starting points
Small cells originated about 10 years ago as residential femtocells, very small, compact and low cost devices primarily addressing poor in-building coverage problems for domestic and SOHO business use. Over time, the capacity and capability of small cells has increased so they can serve large tower blocks, offices and malls with very high traffic density cost effectively.
DAS (Distributed Antenna Systems) have a longer heritage addressing mostly larger properties such as shopping malls, stadiums and transport hubs with high footfall. Off-air DAS systems, which don’t have their own basestations but instead act more like a repeater to improve coverage, have been used for medium sized buildings.
Major Drawbacks
Small cells have a reputation for being single operator only. They must be fully approved and integrated into each mobile network operator’s back office systems.
DAS have a reputation for being costly, based on the difficulty of installing suitable cabling which previously could have involved thick co-ax cables, the full cost of macrocell basestations to drive them and the comprehensive professional services required to plan, install and commission each system.
Addressing the issues
We’ve reported of a couple of system integrators who have simply installed two, three or even four sets of small cells in their customers’ premises to enable full multi-operator service. This involves more radio heads (small cells) than would be the case for an equivalent DAS solution, hence more cabling although this is all Cat5. A benefit is that there can be separate RF plans, one for each network operator, which would take into account the differing coverage and capacity needs. I’m told the cost of the overall solution is lower than deploying a full DAS system and the associated macrocells to drive it.
A further downside of this approach is where both 3G and 4G are required. For operators with approved multi-mode 3G/4G small cell systems, a single radio node can serve both needs. Otherwise an additional duplicate set of small cells would be required throughout – potentially up to eight boxes per coverage area.
Nonetheless, this solution is available and approved for deployment today, solving the problem quickly and effectively.
DAS vendors have come up with a variety of simpler, lower cost solutions. Rather than co-axial cable, we’ve seen some using CAT5 or CAT6 Ethernet cable for the last section and other higher capacity solutions with fibre to each node. CommScope, Cobham, Zinwave are typical examples of systems that act as a relatively straightforward RF pipe from their input distribution point to each node. For much smaller buildings, both Nextivity and Zyxel have launched their own four RF node DAS systems designed for buildings with lower capacity needs.
Driving DAS with Small Cells
A significant part of the total system cost of DAS is that of the macrocells used to drive it. There has been little incentive for macrocell vendors to introduce lower cost products up to now – I’ve heard Ericsson executives explain at public conferences that the number of units involved compared with their global macrocell shipments is very small and that this extra investment would only result in smaller revenues (and profits).
Operators have also told me that there can be some operational issues in managing a few cellsites which have unusual configuration parameters and low power operation. There are also concerns about introducing new radio equipment vendors into their networks, which involves a lot of backoffice integration for everything from configuration, operational monitoring, performance management to spares.
This has led to really quite inefficient solution designs, where large racks of full power macrocells are connected via attenuators into the DAS systems.
A first step is to use small cells to provide the RF signal into the DAS system. These are typically lower cost than the full macrocell equipment. Some of the more advanced features may not be required or appropriate. While some small cell products may not support the very large number of simultaneous active calls/data sessions, they often can achieve the full throughput of data capacity – effectively sharing the resource between fewer people and thus providing each with a faster and better service.
This approach is already in use today where small cells have been approved by an operator, specifically I’ve heard that SpiderCloud is being used by a couple of very large operators as their default equipment choice.
Sharing Spectrum
A more elegant technical solution proposed by ip.access is to share the single frequency of a standard small cell between all operators, using existing roaming standards through a neutral host. Their SUMO architecture would appear to be the most cost effective all round but there is reluctance by operators to allow others to use their spectrum.
Integrated DAS with built-in Small Cells
I’ve recently seen announcements of two products which can operate as Small Cells and/or DAS within the same solution. This is quite unusual because approval of a DAS system is relatively easy compared to that for a new small cell or RAN vendor – DAS appears to the operator as “just another antenna”, whereas introducing a new RAN equipment vendor is quite a big deal.
SOLID, a South Korean DAS vendor added a rack with “signal sources” which can be plugged into their existing active DAS system. While they weren’t keen to brand these as Small Cells, they are effectively a rack of LTE small cell equipment and could be assigned to different operators, or to different frequency bands or both.
Huawei, which has already seen a lot of success with its Lampsite 2.0 product, announced a version 3.0 at MWC this year. It was already possible to separately associate the independent radios at each radio head with a different network operator, using the same physical equipment to serve two operators. But the limitation had been that the operator had already approved and installed Huawei as one of their existing RAN vendors.
One of the big differences with Lampsite 3.0 is that it can also accept RF inputs from external basestations, so you could connect an Ericsson or Nokia (or third party Small Cell). Each RF node is capable of sharing up to 240MHz of RF bandwidth across up to four bands, easily catering for multi-operator, multi-mode and multi-band.
This combination of acting as a DAS system, which requires relatively little certification/approval by operators , alongside built-in basestation capabilities for those already using Huawei RAN already offers a quicker route to market, retaining choice to use other OEM equipment (including third party small cell vendors).
CBRS and higher frequencies
Typically today’s DAS systems don’t support frequencies above 3GHz or so. They could be extended to include the CBRS band at 3.5GHz or higher if there is market potential. While it’s technically possible that DAS could even be used at 5GHz for LAA or MulteFire, there is less value in doing so because the range of 5GHz is much shorter (so would need many more radio heads) and each sector (which can encompass multiple radio heads) would need to avoid conflict with any Wi-Fi access transmitter (access point or end user device) throughout each entire sector/zone.
I would think a CBRS solution using small cells connected via a neutral host should be much more appropriate solution.
Conclusion
There is clearly a growing demand for multi-operator in-building solutions, with likely different approaches for smaller versus larger buildings that will also depend on regional preferences and operator approval.
The most successful will be cost effective, easy to install, require limited approval yet offer full multi-operator capability.
From a technical/architectural perspective, I’d argue that simple independent small cells would be the most cost effective, sharing common frequency band(s) using neutral host (such as ip.access SUMO architectural). Politically this remains quite difficult to gain mindshare, but could become popular for CBRS and/or other shared frequencies.
Parallel sets of small cells appear to be a very practical albeit tactical solution available for immediate deployment. More integrated formats with multiple independent small cells in a single unit could optimise this approach but would be dependent on interest and approval/certification by network operators.
Small Cells driving DAS offers good medium term potential especially for small to medium sized properties or where less capacity is required (eg residential tower blocks). The architecture may become less appropriate in the longer term for LAA or CBRS, although DAS vendors are indicating they may support some of these higher bands.
Integrated small cell and DAS solutions could be attractive for larger properties, but will need approval by all operators.
