Cambridge Wireless, now branded as CW, is a UK industry not-for-profit that organises many technical seminars and events on all aspects of wireless communications. Its Small Cell Special Interest Group is the second most popular, and this event was again over-subscribed.
This session explored how more open access to spectrum could develop provision of cellular service by new players, introducing new funding streams, enabling new vendors and improving all round service quality levels.
Rather than provide a blow-by-blow recap of each speaker presentation, I’ve picked out a few key themes for specific attention.
Operators are outsourcing ever more of their operations
Nick Johnson, ip.access, has been an enthusiast of the neutral host business model for some time. He notes that over recent years, operators have been outsourcing ever more of their operations. It started with sharing cell towers, and has moved across into shared RAN equipment and completely outsourced network operation. Surely spectrum sharing is simply the next logical step.
There are some visible signs of progress towards this approach. Examples of established shared RAN organisations include MBNL UK serving both BT/EE & Three and also CETIN in the Czech Republic serving both O2 and T-Mobile. Mexico has taken the unusual step of regulating for a new shared wholesale-only 4G network, with Altran being awarded the contract last month. More recently Unica Brazil announced the launch of “Small Cell as a Service” using ip.access Viper neutral host platform.
Where operators are prepared to share spectrum through neutral hosts, this would simplify and substantially reduce the cost of deployment. It doesn’t necessarily need new standards or spectrum – 3GPP standards exist today which allow a single set of small cells in a building to offer service connected to all network operators. What’s needed is a further evolution of the business model, where operators delegate operational control to trusted third parties in return for access to in-building cellular capacity.
MulteFire – the new DAS?
To be fair, this wasn’t necessarily the official position of the MulteFire Alliance, but a view shared by many delegates. The technology could provide full LTE service shared by all operators in an efficient and cost-effective manner.
Dirk Lindemeier, a senior Nokia executive and MulteFire spokesman, outlined the three operational modes of MulteFire:
- As a PLMN extension, effectively just another RAN at a different frequency to other LTE bands, directly controlled by an operator.
- Operated by a neutral host, which would require some core network capabilities and a roaming interface with other major network operators.
- Standalone, providing a private network such as for large industrial campuses such as oil refineries, factory complexes, offices etc.
The MulteFire Alliance expects to publish its specification during Q1 2017. Timescale for commercial availability will be mostly driven by handsets.
Read a full explanation of these different modes in our popular white paper on MulteFire.
Spectrum assignment is no longer binary
Simon Saunders of Google noted that up to now cellular spectrum has generally been either highly priced and awarded on an exclusive national basis to large operators or completely free and open for unlicensed use. Dynamic spectrum allocation, as being pioneered through the CBRS scheme in the USA, introduces a sliding scale that includes priority and general access at a reasonable price.
CBRS is less about generating a new tax raising revenue stream for the government (although priority access licences will do that) and more about ensuring the most efficient use of spectrum for all parties.
Federico Boccardi, spectrum expert from Ofcom, confirmed this view. Although LSA (Licensed Shared Access) in Europe has some differences with CBRS, their intention is less about making additional revenue and much more about enabling the most efficient use of available spectrum.
DECT Guard Band Sharing
While there has been a lot of excitement about 3.5GHz and 5GHz bands, the UK also has a scheme to share low power use of the DECT guard band at 1800MHz. A major advantage is that this band is inherently part of Band 3 and widely supported by existing handsets. The band is unlicensed for shared use in Netherlands and Sweden, but has been auctioned off for high power use in some other countries. Ofcom has just completed the comment period for a consultation about what to do with this potentially valuable resource.
3GPP specifications allow carrier aggregation between this 1800MHz band and 5GHz, allowing high speed on demand using LAA. I understand this would typically aim to combine up to three 20MHz sub-channels at 5GHz thus providing the potential for several hundreds of Mbps using Carrier Aggregation.
Ofcom has “outsourced” how best this 1800MHz guard band resource would be shared efficiently as something to be agreed directly between the licence holders themselves.
3GPP formally assigns Band numbers for CBRS, LAA and MulteFire spectrum
3GPP had within the previous few days formally approved Band 48 for the US CBRS spectrum (3550 to 3700MHz), overlapping with Bands 42 and 43 already in place (3.4 to 3.6GHz and 3.6 to 3.8GHz). This was required for both smartphone and small cell vendors to proceed with product development.
Band 46 had already been assigned for the 5GHz unlicensed band 5150 to 5925 (mostly used by Wi-Fi), for use by LTE-U or MulteFire. Not all of this huge large band is fully available for use in every country. Ofcom is actively consulting on how more of it can be released for widespread use.
Small Cell patents held between the large OEMs and several small cell startups
EIP, who kindly hosted the event, shared highlights from their patent database. It seems that "essential patents" (i.e. you couldn't operate a small cell that complies with 3GPP standards without using them) are owned by the major equipment vendors (Huawei, Ericsson, Nokia), chipset vendors (Qualcomm, Intel, Samsung) and several small cell vendors (ip.access, SpiderCloud etc.). There are a few with specialist companies that specialise in owning IPR rather than making products. Some of the acquisitions (Intel indirectly buying Picochip, Nokia acquiring Alcatel-Lucent etc.) have consolidated these holdings.
It was a surprise to me that over 70% of patent filings are granted. I had thought the figure was much lower. Perhaps methods of searching for related patents (called "prior art") are much better than before. Either that, or the threshold to qualify for patent grant has been lowered.The number of small cell patents filed peaked around 2008 when they were first introduced. Many of the relevant small cell features are now integral to LTE and so may not be highlighted as such.
Larger companies negotiate and agree licensing deals on a fair and reasonable basis, but there are occasional surprises when some unknown IPR company pops up and asks for a settlement.
Presentation slides from most speakers are available without restriction on the CW website
[ThinkSmallCell is a CW Small Cell SIG champion, voluntarily organising and planning these events].