The size of the Small Cells Americas event has doubled again from last year, attracting around 500 participants. The new venue was well suited to cope with the influx, with a large and busy exhibition area, two large side-by-side auditoriums and plenty of nearby private conference rooms.
There was a real sense of momentum. Exhibits were no longer primarily of small cell boxes in different form factors. We had a broad range of stands with everything from robust casings, software tools, services and software alongside the latest generation of small cell products. I also hear that the supply chain is seeing greater activity, especially for residential (where Free France is shipping in volume) and enterprise. This all bodes well for a successful 2014.
There were a few surprises though. Snow and ice isn't what you expect to find in Dallas at any time of year, and an unusually strong ice storm caused widespread disruptions to many return journeys home.
Small Cell Forum Release Two
The Small Cell Forum's Release Two publication coincided with the start of the event. Comprising over 20 new and updated documents, it looks to be quite comprehensive, addressing most aspects of Enterprise small cell deployment. These are all freely available to download from scf.io and we recommend you start with the overview document. We also published an interview with Gordon Mansfield (Chair of the Small Cell Forum) last week where he shared some insights around the Release program and the Forum's future plans.
Introducing the event, Gordon pointed out that while North America leads the deployment of residential Femtocells, Asia-Pac leads for non-residential. Surveys emphasise how poor in-building coverage can be (39-61% dissatisfied with it), yet 80% of cellular traffic is now consumed in-building. Traditional DAS solutions are considered far too expensive, whereas small cells provide a compelling business case to both the enterprise and the operator. Payback periods can be very short and often less than a year.
Several delegates commented on how the release program has brought focus to the Forum, which has worked hard to meet the short deadlines set. There is a renewed sense of purpose and achievement.
AT&T scaling up for mass Small Cell deployment
Bill Hogg, AT&T's SVP for Network Planning and Engineering, draws a line at 30 storey size buildings – anything taller than that probably justifies a full DAS implementation, whereas anything less can be addressed through small cell solutions. He expects that crossover point to shift upwards as small cell solutions become more mature.
AT&T have clearly being working hard behind the scenes to put the procedures and processes in place for their Project Velocity, which will deploy some 40,000 non-residential small cells by the end of 2015. This is split between enterprise and outdoor urban (my guess is 75%/25% enterprise/outdoor). Traditional approaches for macrocell rollout simply don't scale up to handle 10,000's of units, so they've updated their software systems and internal processes to handle that. Previously, they didn't have the capability to specify a backhaul link for a light-pole or piece of street furniture – only for a zip-coded street address.
The planning tools they are using to identify precise locations for small cell hotspots are getting better in the X and Y co-ordinates, but the Z-plane (height) accuracy is still pretty much non-existent. I also saw Amdocs demonstrate a smartphone app, which would allow field staff to co-ordinate their plans and progress with back-office workforce management systems.
These changes have allowed them to drive down the deployment cycle time (planned to live operation) to less than 30 days for non-residential. But some outdoor locations remain challenging – Palo Alto's zoning constraints mean you still can be looking at anything up to 3 years!
Softbank Japan continues to densify their network and believe they need three layers – macro, pico and small cells. They will be launching a residential LTE small cell early in 2014. They currently run both FDD-LTE and TD-LTE networks.
China Unicom explained their small cell strategy and test results. They have trialled several 2 Watt 3G small cells to great effect, reaching 1700% data rate increases outdoor and 400% indoor. They have just started 4G small cell trials and are not using a 4G small cell gateway at this stage due to the low numbers of equipment involved.
We also heard about small cell trials and plans from Mexican operator Iusacell and Brazilian Oi, showing this is not just a North American solution.
One of the surprises for me in the Enterprise Release is the solution architecture which now clearly includes the option for some local control within the building. This is targeted at the medium to large installations, and provides both services and signalling optimisation.
SpiderCloud have been offering this for some time, with their proprietary E-RAN solution, which has recently been extended to handle LTE. Quortus have just launched a software subsystem aimed at solution providers who want to develop their own version.
Cisco have achieved a significant milestone, with the 3G upgrade module for their existing Enterprise Wi-Fi hotspots now an orderable part. The full might of the Cisco Enterprise Sales Force can now swing into action. The module is based on Ubiquisys technology and I hear that the acquisition/integration is going well, with several key Ubiquisys personnel now assigned leading Cisco engineering and marketing roles.
Although not visible at the conference, I also understand that Verizon have been busy with many 3G CDMA enterprise small cell deployments (using Ubee-Airwalk equipment). So it isn't purely a 3G UMTS story. Although some 4G only enterprise small cells were on show, there didn't seem to be an appetite from them in this region – the 3G service is still needed to address indoor voice coverage for existing handsets/smartphones.
Athena Wireless have combined a 60GHz backhaul with their own 4G metrocell, called the Pixie. This makes for quite a compact physical package and is one of the few integrated metrocell/backhaul products I've seen. (Airspan has had a 4G metrocell within integral NLoS backhaul for some years).
Two Non-Line-of-Sight backhaul vendors were reporting considerable progress over the last year. Fastback Networks and Tarana have conducted successful field trials, demonstrating 100 to 500 Mbps over short range (up to 1km) in hostile environments such as downtown Manhattan. Both use Qualcomm's silicon (based on the DesignArts acquisition), with FastBack focussing more on using unlicensed spectrum at 5.8GHz.
RADWIN, with more pedigree in Non Line-of-Sight, presented a series of case studies showing that the most obvious direct path isn't always the best. Some specialist knowledge is needed to get the best performance and reliability out of the technology.
I wasn't able to see many of the other backhaul presentations which ran at the same time as the main conference sessions.
Metrocells do work
Rob Soni, CTO of Small Cells for Alcatel-Lucent, was keen to point out that outdoor 3G small cells really do work, even when using shared frequencies with the macrocells. He ran through some of the technical issues which crop up and explained how they had been overcome. In all deployments to date, they'd witnessed higher total system capacity and higher end-user data rates. Alcatel-Lucent launched its Metrocell Express Site Certificiation program - a database which qualifies and publishes a list of candidate outdoor metrocell locations containing over 500,000 entries from their partners already.
Looking ahead at 4G small cells, Rob described an Extended X2 interface (eX2) which passes enhanced messages between the small cells and macrocells to co-ordinate interference mitigatation and improve performance. Faster updates can provide better results – Qualcomm studies show that updating the ABS (Almost Blank Subframes) schemes every second rather than every 15 minutes can provide efficiency improvements of 30% rather than 15%.
Rural Small Cells
Contela won my award for stretching the boundaries of the Small Cell terminology, launching a 20 Watt RF small cell for rural applications. 20 Watts is the maximum regulatory transmit power from a macrocell, and few would have considered anything more than 1 or 2 Watts to be a Small Cell in the past. They described several remote/rural applications.
Several of the rural Tier 3 telcos described where Small Cells might fit their needs. Unlike other countries, national roaming is commonly used in the US to access service in remote areas and Tier 3's get paid for providing coverage in places where the Tier 1 operators don't reach. These can be anything from the truck stops (where many trucks park overnight and watch NetFlix all evening), to the insides of shops/buildings which aren't penetrated by remote macrocells. Examples were given where a $10-15K revenue stream is achievable by deploying a Small Cell at these types of locations.
Satellite backhaul was also mentioned – a domestic $50/month satellite service is enough to support a rural residential femtocell serving one or more adjacent buildings.
ClearSky and Cellcom both offer a Small Cell managed service, removing the high entry costs of initial small cell deployments for these smaller Tier 3 operators. Some rural areas still have such poor coverage that some retail cellular stores don't have any cellular service – staff have to walk out to the street to find a signal in order to activate phone service for their new customers.
Weatherproof toughened casings are commonly required in the US, especially where locals literally take pot-shots with their shotguns at telecoms equipment. Cable Labs have a specification for shotgun proof enclosures and Lindsay Broadband were exhibiting theirs. These could be strung on a street cable, allowing a cable TV company to quickly install and service an area. Cable companies have access rights to the locations (street wiring), power (up to 200 Watts) and backhaul (cable broadband) which would allow them to deploy quickly and easily.
For added protection, perhaps they should be running "bullet proof software" too.
A Radical View Ahead
Nick Johnson, CTO of ip.access, gave an amusing and challenging presentation. He argued that the value of licenced spectrum is reducing towards zero, and that the capabilities of Wi-Fi are evolving to match LTE-Advanced. The provision of services (such as voice telephony) and access (whether cellular or Wi-Fi) are becoming more distinct. Perhaps operators should consider turning off Rich Communication Services and leave these to the Over the Top providers such as Google and focus on providing a really good connection experience throughout?
His use of a Monty Python comedy sketch added an amusing dimension to the more serious topic of the operator's future business strategy dilemma.
- Internationl roaming is getting much cheaper. My UK Network provider "3" has removed all roaming charges to the US, and allows use of my bundled minutes/data allowance without extra charge. T-Mobile USA also offers similar free roaming to over 100 countries at no extra cost. Both have common sense usage limits, and are not intended for extensive video streaming but would support email and basic web browsing. The hotel won't be getting $13/day for my Wi-Fi service next year for sure.
- iPosi was demonstrating their supersensitive GNSS receiver in the depths of the hotel building. This not only picks up a signal virtually anywhere indoors (receiver sensitivity is down below -170dbM) but is less susceptible to jamming.
- Quortus launched their Small Cell Enterprise local controller software, meeting the new Enterprise architecture from the Small Cell Forum. The software can even run "in the cloud" in situations where it isn't appropriate to deploy a dedicated local server onsite. It's aimed at systems integrators and solution vendors, who can customise and adapt where needed.
- Competition between Small Cell chipset vendors remains fierce. I'd say we are down to no more than six contenders at the moment (in alphabetical order): Broadcom, Cavium, Freescale, Mindspeed, Texas Instruments & Qualcomm. The wireless assets of Mindspeed are very likely to be sold and an announcement is anticipated imminently.
- Several LTE only small cells were on display. Fujitsu (Japan) and Aritel (Korea, supplying the LG+ network) both want to expand their market outside Japan, as do Qucell and Contela (Korea). Purewave had their dual mode 3G/LTE box on show, and said they'd made progress with several LTE contracts but (sadly) couldn't publicly give details. Athena Wireless were showing their Pixie box, which can also be integrated with their 60GHz backhaul.
- Free France, which we've reported about earlier, really is shipping substantial numbers of residential femtocell modules now. I'm told it might not quite have reached 100K a month yet, but certainly many 10Ks. At this rate, they will quickly become the largest non-US femtocell network.
- Radisys reconfirm that the most popular LTE-Advanced feature is Carrier Aggregation. They've seen greater diversity of suppliers for 4G small cells than for 3G, with Korea being a major proving ground.
- u-Blox launched a new high specification GPS/GNSS receiver and oscillator module aimed at the small cell market.
As usual, my apologies for omitting many other aspects from the conference - you'll just have to attend if you want to learn more yourself. Avren/Clarion did a super job of organising the conference and handling the inevitable glitches seamlessly. Many people remarked how the size and scope of the event allowed plenty of networking and meetings to be scheduled.
I'm just pleased to have resisted the 26 ounce steak on offer in the restaurant. It looked enormous!
We also captured a sense of the event through several short interviews - about 9 minutes of video in total uploaded to YouTube including ip.access, Radisys, Quortus, FastBack and Cisco.