As the great and good assembled for this year’s Small Cell World Summit, there seemed to be a little despondency which contrasted with the recent upbeat WBA Wi-Fi event. Was this because attendance didn’t grow as before, the somewhat impersonal/industrial venue, the clash with IEEE event next door, the somewhat unstructured/unbalanced conference agenda, a wider and more complex scope, or simply because the industry has moved on to the next stage of maturity?
A theme picked up by several speakers was that perhaps the industry has been accused of “Crying Wolf” too many times in the past, and larger signs of progress will be needed to convince the more sceptical. Mansoor Hanif of EE thought that "the Wolf is starting to feel hungry now" citing the lack of dark fibre as one reason to switch more investment across from macro to small cell.
The regulars were all in attendance and many told me of worthwhile meetings. The exhibition area was well set out with plenty of interesting demos to catch the eye.
The residential market segment has been written off by some who view Voice-over-Wi-Fi as the inevitable solution (complemented by smart boosters where wireline broadband isn’t available). EE UK is rolling this out at present. Mansoor Hanif, Director of RAN Development and Programs, was quite forthright in spelling out their intent. It’s available for use where no other option exists (i.e. outside any other type of coverage). They plan to ramp up shipments of 3G femtocells (a figure of several hundred thousand before end 2015 was mentioned), and I estimate this could pull them ahead of other UK networks next year in terms of volume deployed. My guess is that this is pre-empting strong competition from either BT or TalkTalk who have been trialling prior to mass deployment. It's hard to guess if or when BT might roll these out (a competitor pointed out the combined BT/EE would likely have to give up some of its precious spectrum portfolio during the acquisition process. Once the status is known, they'll both rethink their strategy).
When asked about whether there was concern about reliability and quality of VoWiFi, he said (I paraphrase) “The best feature of Voice over Wi-Fi is that you can turn it off” which has to be explained to customers who might otherwise get a bad experience.
Joe Madden of Mobile Experts, predicts continuing growth in this sector with a one-off boost this year. I’d expect this is partly due to Free France updating all their set top boxes (said to be 1 million units alone), and the potential for one or more big launches in the short to medium term. His chart below is published by the Small Cell Forum and this year's figures are based on known orders for parts from the component suppliers.
The context is that Wi-Fi and cellular both have a role to play in the home; neither will win out completely. Wireline broadband players, such as Free or TalkTalk, could make most difference when allowed access to even a small sliver of spectrum and/or MVNO wholesale roaming. You can detect some growing interest in 4G residential, but it’s predicated on VoLTE being deployed first. Contela were pleased to be exporting their 4G small cells into Japan; it's been difficult for the Koreans to win deals outside their home base despite having proven and mature solutions, but their persistence is paying off.
LTE small cell gateways are recognised as essential for large scale deployments. Zyxel had theirs on display, which is "trial ready", built using commodity x86 based hardware and aimed at the smaller operator. Multi-mode products were also on show, such as Alcatel-Lucent's 3G/LTE residential 9961 based on the Qualcomm chipset.
The forecasts show this to be the most burgeoning sector, with up to 450K units shipping this year. The most exciting segment is the “Middleprise” large enterprise and venues. Cisco were showing off the Spidercloud plug-in module for their Aironet Wi-Fi access points which is now in production. I understand a lot of training and behind-the-scenes preparation has gone on to bring this to market. What’s different to their Ubiquisys offering is that the size of each deal is much larger. Bluntly, the sales bonus makes this more attractive to salespeople and worthwhile overcoming any resistance from the network operators or minor practical issues. Vodafone presented a series of successful case studies of Spidercloud deployments (which they brand as Sure Signal Premium), saying they were deployed “in a Wi-Fi like manner”. Cisco told me they were frequently getting many calls from Enterprise IT people who want to deploy their small cells but are frustrated that they can’t connect them to a network operator. Informally, another operator told me they weren't against such connections in principle but it just doesn't seem to be their business priority at the moment.
Airvana’s One-Cell also had a strong advocate in the form of Nex-Tech Wireless, a US Tier 3 operator in Kansas. They installed a system in 75 days (concept to live use) in a 7,500 seater 322,000 sq feet multi-purpose stadium. It was half the price of upgrading their existing 3G DAS (which was left in place), and much less complex than deploying Remote Radio Heads. In any case, the nearby macrocells are already fully loaded serving the University Campus next door. Dual band 4G at 1900 and 700 was augmented by single band radio heads in less demanding parts of the venue. The physical install took only three days, a timeframe I’ve heard quoted by Spidercloud in the past and frequently derided by those used to traditional methods.
Huawei walked through a case study of a Lampsite deployment at a major railway station in China. It took 14 days to install (including the configuration/commissioning) and could be enabled for additional capacity later through software configuration and licencing. An interesting point was that this wasn’t just about data offload from the macrocell. While that’s a good step, the improved user experience can change user behaviour – if people see a better service, they’ll use it more and that results in more revenue. Their ROI (Return on Investment) should be less than 2 years, and can be even less than one year if there is faster unforeseen increase in data usage.
SOLID revisited the view that there’s room in the toolbox for both DAS and Small Cells, revalidated by a panel session on the topic. But operators thought the scope for small cells is growing – AT&T suggested that buildings up to at least 400,000 square feet might be appropriately served by small cells. We were also reminded that the DAS industry continues to develop and is bringing out lower cost solutions.
Scott Morrison, VP/GM of Cisco's Small Cell business, spoke of their vision of "Smart Workplaces" where Wi-Fi and cellular are seamlessly blended together and enhanced by a range of services. A similar angle was demoed on the Intel stand, where ACS have put together a mobile edge computing demo with several enterprise apps and features. It's a theme that has evolved over time and builds on the foundations of good inbuilding service and low cost, powerful servers. They appear to have a growing supplier relationship with EE UK.
Share and share alike
Dean Fresonke of Clearsky, who offer a neutral host commercial managed small cell service told me that most building owners want a single solution for all operators (and a single point to call for support). But in the US at least, each operator wants their own spectrum, presence and control so this needs multiple small cells. Their solution consolidates and combines to satisfy both, and he’s still of the view that multiple small cell installations are cheaper than DAS.
Nick Johnson of ip.access also believes in sharing, and proposed wider use of the MOCN standard between networks. In another aspect of sharing, they’ve teamed up with Amdocs, iBwave and others to create a HetNet Ecosystem of managed services and certified partners to accelerate rollout.
Urban Small Cells
If more traffic is offloaded into buildings using these Enterprise Small Cells it would delay the need for outdoor urban because the capacity and scope of macrocells could meet demand. One European operator pretty much confirmed to me that that was their expectation.
However there are rumours of activity in other regions. China has had huge investment in LTE macrocells of late (650,000 macrocells to date) and is said to be moving into an urban infill deployment phase as well as in-building during 2H2015. Reliance Jio of India is rolling out a huge LTE only network supplemented by Airspan small cells, powered by Qualcomm chips.
JCDecaux (who won a Small Cell Forum award) were demonstrating how an urban small cell can be unobtrusively integrated into a standard advertising outdoor display.
Nokia told me their urban Flexi-Zone 4G cell has huge capacity, possibly even more than a macrocell sector, with which they’ve had a lot of success. In this market segment, it’s less about numbers deployed and more about the impact they make to offload the macrocells nearby. An ATT presentation showed how relatively few (tens of installations) could release huge capacity from the existing network.
Rural and Remote
Forecast figures showed that this sector would numerically exceed the urban market, with some 20,000 shipping this year. One vendor thought this was an underestimate.
EE have committed to expanding rural coverage throughout many parts of the UK and have been actively working with Parallel Wireless. A presentation explained their target of 1,500 locations (communities) by end 2017, each of which may have several small cells interconnected by a mesh, so the total volume could be 5,000 cells or more.
One new vendor exhibiting a brand new Small Cell design at the show was Deltenna, a UK company, which had built a compact standalone 4G outdoor unit called LittleWolf powerd by a Cavium chip. It’s intended for use as a relay, combining a receiver (typically with a high efficiency directional antenna to the nearest macrocell) and LTE small cell capable of operating in any of four bands. The unit only requires 20 Watts of power which is quite feasible from solar panel and battery. RF output power is limited at this stage but adequate for a neighbourhood.
What’s more is that it has the Quortus core network software onboard and can act as a completely standalone network. Quortus ECX Tactical has some quite amazing features that allow it to morph between standalone mode, pair with nearby units or mesh into the existing network – all without dropping calls/sessions. Install these on every fire truck or police car to bring your own network with you – as someone pointed out, the first thing the fire brigade might do when arriving onsite at a burning building is turn the power off. With the UK expected to switch off its old expensive TETRA network and migrate to sharing the commercial LTE system, this could be quite a robust and complementary capability.
I’ve not radically changed my stance on the market segmentation of small cell chipset vendors, published last month. Intel and Cavium would both point to wins in standalone small cells, especially TD-LTE for Intel (using the earlier Mindspeed Trancede design as pictured below) and the higher capacity LTE units. But their main focus is in making a play for RAN virtualisation and Cloud RAN architectures, with Intel pushing hard on Mobile Edge Computing.
Qualcomm are clearly making progress with their FSM99xx, now incorporated into Airvana’s standalone product, Airspan’s LTE product line, ip.access Presence-cell and Alcatel-Lucent’s multi-mode 3G/LTE residential and enterprise products on show.
Broadcom are quietly continuing to gain market share, having dominated new residential and enterprise designs for some time. They told me their main differentiation is starting with a low power architecture and scaling up. They’ve integrated the RF to reduce cost but enable lots of hooks/APIs to allow third parties to innovate. Their sniffing technology allows constant sampling of real-world conditions and they told me “We don’t make the SON but we make the SON better”. Publicly, they’ve reported to have shipped 2 million units and privately they’re doubling the pace. They don’t imagine there is any reason to do anything differently as a consequence of being acquired by AVAGO.
There were a few private demos going on behind closed doors, but broadly speaking there are now a number of mature chipset reference designs (and software stacks) available in production today that provide all the 3G/LTE features you need including TD-LTE, Carrier Aggregation, up to 20MHz bandwidth, multi-band and MIMO.
Small Cell Forum Awards
The awards dinner this year was held at the Greenwich Maritime Museum, sited on the prime meridian (zero degree longitude) agreed in 1884 (from almost 40 choices) and the Longitude Prize to discover accurate marine timeclocks originated in 1714. It was a suitably distinguished setting for the Small Cell Forum awards. These are highly prized by the winners, who very much deserved them – as a judge myself, I can confirm the judging process was quite involved and diligent - and all including those who were shortlisted are noteworthy.
Pictured below are Sue Monahan CEO Small Cell Forum introducing the event while Andy Germano, Vice President of the Small Cell Forum, prepared once again to amuse the crowd and looking worried he's misplaced the secret supply of jokes hidden in his coat pocket.
While the show lacked some of the buzz of previous years, there are definite signs of steady progress. Voice over Wi-Fi isn’t about to kill the market for residential Femtocells (why else would an operator actively invest in both). The Enterprise sector (especially the medium to larger end) looks very likely to burst into life, possibly at the short term expense of some urban deployments. Rural and remote are at long last getting some attention.
Perhaps it’s because the telecoms industry has been overdriven by the hype of new technology for so many years, overly keen to learn about the latest mind-boggling physics-defying innovation. Customers don't want peak speeds, they want consistency, reliability and availability everywhere at a sensible price. While EE announce 300Mbps capability throughout London, it's their investments in raising the bar by reaching rural, residential and enterprise that should make the most difference to customer perception.
I’ve come away rather more optimistic and almost disappointed there wasn’t more hype during the week. Perhaps we need a bit more active conflict on the panel debates, filling them with more contrarian views. Now that we have products with both Cellular and Wi-Fi integrated, it seemed strange to split those tracks – both need to become more complementary rather than competitive with each other. Debating those issues more strongly should help educate us all on the best way forward.
The date for next year's event hasn't been published but I hear it may be brought forward by a couple of weeks or so into May.
(I'll cover the Small Cell Forum LAA workshop that followed after the event in a subsequent report.)
Agree? Disagree? Thought I missed out or misrepresented something? Add your comment below (which can be anonymous).