MWC comprises a huge exhibition, tons of meetings and a co-located conference. Here we review it from a Small Cell perspective. In my view the conference itself was pretty dull (fireworks party without matches). The keynote themes were broadly to expand and grow user base, exploit the SIM card to authenticate for access control in many other ways, and evolve data capacity. The exhibition itself featured $25 Firefox smartphones, an enormous variety of Apps (especially wearables and the Internet of Things), mobile payments, and plenty of network solutions (including many new Small Cell and related aspects). The event seems to be much more about meetings now than anything else, with perhaps only 10% of delegates attending the conference presentations, and remains the number 1 worldwide mobile get-together. The official Small Cell conference track was pretty tame - Vodafone have deployed 300K Small Cells in total, KT (Korea Telecom) and Radisys spoke of 18K LTE deployed in mostly indoor metropolitan areas. Vodafone said they continue to drive vendors to deliver multi-technology small cell and backhaul products with high operational efficiency and look for added value to help the business case. By contrast, the Small Cell Forum booth hosted extensive and popular presentations and is perhaps outgrowing its booth format.
A key network equipment vendor theme was SDN (Software Defined Network) and NFV (Network Function Virtualisation). We can expect next year to see this evolving to orchestration - better methods of managing and manipulating these virtualised software components, but in the short term it means slightly less or cheaper hardware. Frankly, I was more impressed to see Huawei now supporting any of 2G, 3G or LTE (FDD&TDD) on the same physical macrocell radio hardware modules - true software definable radio. We are beginning to see that capability for Small Cells too, but it's not quite as mature yet.
Most of the Small Cell activity is around 3G indoor (Enterprise) and LTE outdoor (Urban), with 3G still important indoors (for voice) and LTE HetNets seen as the longer term solution for capacity. At least four DAS vendors announced lower cost, simpler products intended to address larger buildings and stadia - highlighting the growing demand for in-building cellular solutions. Many new LTE Small Cell vendors are appearing on the scene. Residential femtocells still have a place in the market especially where integrated into a broadband modem or set-top box, driven by a different business case than before. There were some signs that the radical approach of Free France, who are shipping many 10Ks of femtocells a month, may be emulated by others.
It's becoming even clearer that there is a need for in-building non-residential solutions. 80% of connections are made indoors but only 2% of commercial buildings is covered from inside.
It's not just about adding capacity, but ensuring consistent quality and comprehensive service. There are now a wide range of technical solutions, some announced, some in trials, some proven and a few commercially available. Several DAS vendors launched lower cost solutions, primarily addressing the high end of the market. The consensus recognises that a combination of Small Cells, DAS and Wi-Fi will be required, but may disagree about where the dividing lines are drawn. Gordon Mansfield of AT&T told me he thought the line between DAS and Small Cells would shift up and down as vendors respond to market demand.
Specific new DAS products announced this month included:
- Commscope (who had acquired Andrew): ION-E. Pumps an astonishing 300MHz of any useful cellular spectrum bands down a CAT6A cable to a set of ceiling mounted radio heads. Also carries IP for use by associated separate Wi-Fi access points. Handles a mix of any bands up to 2.7GHz. This is the "heavy artillery" to use where you can justify multi Gbps, multi operator in a single room or stadium.
- Kathrein: K-BOW. Positioned as a sort of "mini Cloud RAN", this backhauls all the radio signals over dark fibre, potentially to an offsite central location which could host multiple remote buildings such as a hotel chain. Can selectively switch/redirect capacity to different parts of the building at different times of day.
- Axell: idDAS . Focusses on more actively managing the total macrocell capacity by dynamically redirecting it to different parts of the buildings.
- Alcatel-Lucent/TE Connectivity: Announced a direct integration between the macrocell vendor's basestation and TE Connectivity DAS systems using a CPRI interface, effectively bypassing some of the expensive RF bits. This provides an architecture similar to the Ericsson Radio DOT or Hauwei Lampsite, but is commercially available now.
The key differentiator between Small Cells and DAS is where the RF baseband processing is performed. Small Cells do this at the local radio head, so only need a low cost/simple IP connection - it can even be a shared VPN from the existing Ethernet network. DAS systems need specially dedicated connections to traditional macrocell base stations, and have needed a lot of RF engineering design/professional services to install and commission. They've previously been tailored to specific frequency bands, and often relied on multi-operator sharing to fund the high cost in $100Ks. These new products should help improve their appeal for specific use cases.
Cisco continue to build on their momentum. Their 3G "snap on" module which quickly upgrades their existing Wi-Fi access points is matched by a similar LTE variant. There's also a new similar form factor unit that can accommodate two cellular modules without Wi-Fi. These could be 3G and/or LTE or even two 3G modules handling different network operators. Cisco told me that CIOs attending their upcoming customer conference have selected inbuilding cellular coverage as the most popular seminar. CIOs are under increasing pressure to facilitate good indoor cellular service for their staff and visitors, and are pressing operators to find ways to meet that need. The construction materials used in new buildings which reflect outdoor RF signals is further exacerbating the situation. You can just Imagine the conversation between the CEO and the CIO when the former can't make quality mobile calls in their expensive new office building. They've also added the Ubiquisys residential femtocell into their portfolio in a nice, compact new form factor snappily entitled the 3331. The SON capabilities from both IntuCell and Ubiquisys have been merged into Quantum SON, proven with a live demo they showed me.
Spidercloud's stand was much busier than last year, showing off their latest 3G/LTE dual mode in-building solution. NEC have upgraded their Small Cell Gateway to dual mode to support this. The solution addresses dense indoor enterprise applications, co-ordinating up to 60 or more radio nodes and resolving a whole new set of challenges around interference mitigation. In addition to Spidercloud's EASY-30 rapid 30 day deployment process, they demonstrated a very simple iPhone App which allows a rough estimate of the number of Small Cells required. For a next level plan, iBwave demonstrated a really easy to use In-Building planning tool (iPad app). Ranplan's inbuilding planning tool was also in evidence at the show - they launched a free licence option for their "lite" version.
Alcatel-Lucent's project with Qualcomm will come to fruition in the second half of 2014, working towards a 3G/4G/Wi-Fi combined in-building product with very impressive specifications. Qualcomm are also embedded in ip.access new 3G Enterprise Small Cell, which are said to enable new types of location and/or presence related services.
The South Korean stand included Contela, who continue to evolve their 3G and LTE portfolio, joined now by Aritel who were showing off their LTE Enterprise Small Cell, deployed by LG+. Operator KT revealed they had deployed 18,000 LTE Small Cells to date (10K in Seoul and 8K in other metropolitan areas).
Tuesday saw the Small Cell Forum issue Release 3: Urban Foundation which contained a surprising number of documents (18). Gordon Mansfield, Chairman of SCF, told me it exceeded his expectations, especially since it followed so closely after December's Enterprise release. The same release co-ordination team have presided over another mammoth edition. The complete Urban document set will be published in June, now renumbered Release 4, and the roadmap for updates and further releases will be finalised in April. The Forum also has a new CEO, Sue Monihan, who brings expertise from running the GSMA North America organisation and I'm sure will continue to expand the remit, membership and influence of the organisation. She will continue to work part time for both organisations concurrently.
Many (but not all) organisations now seem to be focussing on LTE+Wi-Fi rather than 3G/LTE/W-Fi multi-mode for urban capacity solutions - something that Vodafone specifically stated during the conference.
Mike Schabel, VP Small Cells at Alcatel-Lucent, pointed out the radical shift from voice to data is still fairly recent and has had huge impact for operators. Their commercial activity in Small Cell trials and business opportunities has never been busier. Speaking at the conference, he said "I can't keep up" with the number of RFPs today. They are working with 65 operators around the world (without double counting). This is happening in all regions - it's all over the world. While the technologies are the same across the small cell sector, what varies are the different business drivers found across their different customers.
Randy Cox, Head of Small Cells at NSN, showed me their latest LTE+Wi-Fi Urban Small Cell. This is quite a compact unit (about the size of a large dinner plate), capable of up to 400 concurrent sessions/calls. It's very much a downsized macrocell, reusing the same TI silicon chipset and optionally connected through a Flexi-Zone controller. They aren't planning a 3G/LTE multi-mode product at the moment, believing that LTE is the future. "We are in this for the long term" he told me. They continue to support their existing femtocell customers (using NSN's femtocell gateway and Cisco/Ubiquisys small cells), but it seemed to me that their Small Cell focus had shifted firmly to LTE.
NEC bring their own LTE expertise from Japan, with a range of in-house products complemented by partner solutions, both for Small Cells and wireless backhaul. They've selected Radisys as their LTE Small Cell software supplier for their next in-house products. They position themselves as a one-stop shop for the new last mile, offering to handle everything from the initial planning through operational management of a portfolio of equipment. They told me that where macrocell deployments might have had four or five standard templates or scenarios, they've seen more like 400 to 500 different use cases. I do hope that's an exaggeration!
I saw several supporting technologies to de-risk HetNet deployments, ranging from smarter RF planning tools with more Small Cell features, SON vendors making further inroads, test validation and performance loading. There is a wider ecosystem here, finding newer and smarter ways to plan, manage and validate the network which wouldn't be cost effective at scale using traditional methods. Highlights included ERCOM's integration with Infovista/Mentum, allowing emulation of huge numbers of users in real-world scenarios, characterising the behaviour of small and macrocells that can be fed back into the planning tool. IXIA also have a comprehensive test and validation suite, de-risking the rollout of new features or vendor equipment by pushing the envelope and exploring the limits - not just conformance testing against standards. Sources of usage data now include directly from the handset, not just the network. OpenSignal published their "Global State of LTE Report" with actual performance results from 6 million users worldwide. Apparently Sweden has the fastest LTE in the world. The average speed worldwide is now just over 10Mbps. This whole area of back office planning and network management is evolving rapidly with many new technologies, tools and approaches coming on stream. Don't be distracted by the concepts of SDN and NFV - there are many more relevant and beneficial changes to how networks are managed coming along.
Fibre is first choice for Urban Small Cell backhaul, but outside Korea, Japan and a few lucky places, it's not feasible in most cases. ADVA promote their AnyCell fibre solution which combines dark fibre for DAS with simpler IP/Ethernet for Small Cells. Elsewhere wireless has a key part to play in this "last mile". 60GHz V-Band continues to look like a good short to medium term option - sub10 have been shipping their Liberator product for related applications and are well positioned. They announced an E-Band (70/80GHz) product at the show, telling me that they were keeping the pricing options simple with choices of 100Mbps and 1Gbps. Siklu have also been involved in small cell trials using their in-house technology, as have NEC. SyncE and IEEE 1588v2 are mandatory for phase timing, in order to support LTE Advanced features.
Point-to-Multipoint solutions are also part of the backhaul toolkit. CCS have been successful with their mesh based solution, which has a 270 degree line of sight approach and small form factor making this very easy to deploy. They claim 15 minutes by typical lighting contractors. China Mobile are using these in their live LTE network already. Steve Greaves, their CEO, told me they aren't particularly married to any specific frequency band - their unique features are about self-organising and simple installation - and claim total capacity of 480Mbps per hub. Competitor CBNL has just doubled the capacity of their VectaStar product range which now sports up to 600Mbps per hub. John Naylon, their CTO, told me CBNL has successfully expanded their sales activity of late, resulting in rapid growth from LATAM specifically, which has grown from zero to something like 25% of their total revenue in just the past 18 months. (That's not for Small Cell business though - it's many other use cases).
Non-Line-of-Sight. Several companies continue to make headway with products that appear to defy the laws of physics, providing backhaul links under near or non-line of sight conditions at hub data rates up to 400 or 500Mbps. These use many of the techniques of LTE plus some proprietary "secret sauce". Some alignment is required, and due to signal reflections from buildings it may involve pointing other than in the most obvious direction. Blinq has been working on this for several years, and was demonstrating their 2nd generation X-1200 product. This is working its way through the Tier 1 carrier approval process and is well beyond the early trial stage. Fastback Networks target the use of unlicensed bands and announced European compliant product variants. Tarana reports successful trials from some of the most challenging locations in Manhatten. Their 1st generation box is already shipping commercially for use as macrocell backhaul, and their 2nd generation product announced this month is planned for GA in Q4 2014. I also spoke with Max4G, who have also been in NLoS operator trials and report encouraging results with a dual band approach. They've achieved 150 to 170Mbps even in very challenging conditions (peak of 900Mbps), and believe that using licenced spectrum (say 2.6GHz TDD) would be a more predictable and consistent choice than unlicensed (although they could supply either). I understand that spectrum regulations allow use of proprietary schemes within the licenced TD-LTE bands provided the signals don't interfere with other bands/uses, making this an attractive and accessible option.
Almost all of these NLoS products are using Qualcomm's chipsets (including assets acquired through DesignArts), supplemented by additional processing and/or specialist hardware. If the results being claimed by these technology vendors are to be believed, it could capture a significant proportion (but far from all) of wireless small cell backhaul market. They do seem to have proven that Small Cells can be sited in the most challenging (and perhaps also most useful) locations. Next they have to demonstrate this scales efficiently if many were to be deployed in the same dense urban area.
It seems to me that there remains a lot of uncertainty about how much backhaul capacity/bandwidth is required for Small Cells, with some operators seeking to dimension for peak throughput while many others are satisfied with 100Mbps or so per cell. Phase synchronisation (using SyncE and IEEE 1588) is pretty much mandatory, even where GPS is also used.
Ceragon summed up the story for mainstream backhaul vendors to me very succinctly. Pointing at their product portfolio displayed on the wall, it's clear they have solutions for all of the above. About 90% of their revenue today comes from macrocell backhaul - Several billions annually. All of the Small cell technologies contribute only a few hundred million combined. If and when the market takes off, they are positioned to run with it, but are not reliant 100% on it. I've heard this from most of the backhaul vendors at the show, who are cautious about market predictions but keen to be able to "catch the wave" as the opportunies arise.
ZyXel, who have supplied around 500,000 residential femtocells to date demonstrated their latest form factor. The phablet sized unit embeds a 802.11ac Wi-Fi receiver as well as the usual 3G femtocell and Wi-Fi Access point. This allows it to be placed anywhere in the home, using the 802.11ac link as the backhaul to the nearby residential broadband modem.
TIM Brazil announced a deal with Alcatel-Lucent to supply femtocells for Brazil. Its about time - the regulator had removed the need to pay a base station tax which had held back deployment. The agreement covers all categories of 3G Small Cell - residential, enterprise and Urban.
On the grapevine, I also heard that Free France are shipping substantial volumes of femtocells today (many 10Ks/month). All new set top boxes have one as standard. Fewer of their standalone modules are being sold, because customers don't see the value of bothering to install one (even if free) if they already have good voice coverage at home or because they aren't very technically minded. I'm hearing of several organisations who are looking at Free's business model and mode of operation with a view to launching something similar. The key difference from earlier residential femtocell launches would be embedding them within the broadband router or set-top box. As one industry veteran pointed out to me at the show - the industry engineered a really great technical solution for residential femtocells but didn't properly address who was going to pay for it.
One of the best demos I saw at the show was at the DeviceScape booth. They really seem to have cracked the challenges of seamlessly identifying, logging into, validating and switching across to Wi-Fi where sensible (and sticking with cellular where not). Their demo showed the phone logging into a known Wi-Fi hotspot in the background without the usual painful login screens or dropped audio streaming sessions. They've just signed up Virgin Media in the UK and continue to expand their footprint of 20 million users and 20 million known/validated hotspots in what they call their Curated Virtual Network.
The Wireless Broadband Alliance ran their conference in parallel during the event, publishing a document that specifies what they mean by Carrier Wi-Fi. A live demonstration of HotSpot 2.0 showed its potential, powered by Accuris who make other Wi-Fi networks look to the cellular operator just like another roaming partner cellular network. Edgar Figueroa, CEO of the WI-FI Alliance, reminded me that 75% of wireless devices shipped today are Wi-Fi only and won't have SIM cards, but can take advantage of this. Ruckus Wireless pointed out to me that there is quite a difference between carrier class Wi-Fi and lower cost residential Wi-Fi products, and I've seen evidence that non-residential Small Cells are embedding higher quality Wi-Fi modules as standard.
The landscape for Small Cell chipsets continues to evolve in a market which has high stakes (measured in $100 millions) and long cycle times. Qualcomm continue to make progress, with wins at Airspan, ip.access and Alcatel-Lucent. They were demonstrating their Ultra-SON technology that has been actively trialled and evolved at their own campus, optimising the inter-cell handovers and increasing system efficiency. They were also promoting using LTE in the unlicensed spectrum bands - known as LTE-U, an approach that is also supported by Alcatel-Lucent. Their chipset is all encompassing and includes even the RF front end chip which in other designs would typically be bought from ADI or Maxim.
Broadcom also continue to invest heavily in Small Cells, with full 3G and LTE solutions. The demo on their stand showed their LTE system driving three different RF chipset front ends (ADI, Maxim, AKM), with impressive Eb/No signal/noise figures for each. Their chips are included in the majority of Small Cells today, including Alcatel-Lucent, Cisco and many Asian designs.
TI continue to expand their Keystone architecture, releasing yet another chip variant. I understand they remain embedded in NSN's and Cisco's LTE Small Cells, and were publicly announcing selection within PureWave Small Cells and AirSpan's backhaul products.
Cavium announced a new LTE Small Cell platform based on their OCTEON architecture and that Taiwanese vendor Arcadyan had selected them for their LTE products.
Intel, who are acquiring the Picochip/Mindspeed wireless assets, didn't have too much to say this year. Their story should become clearer after the organisational changes have worked through, and may encompass wider technical approach.
I did not see a Small Cell specific announcement from Freescale, the other major chipset vendor in this space, but they had previously announced selection by ZTE for their latest LTE small cell products.
OTHER INTERESTING AND UNUSUAL STUFF
A couple of companies believe they have defeated the laws of physics:
Artemis pCell uses an array of PureWave Small Cells to reuse spectrum, creating a unique signal for each mobile in the target coverage area, and ideal for very dense traffic environments. It does need your mobile to be within coverage of several small cells and cancels out the signal from other users. In a quite separate project, Kumu Networks believe they've cracked the ability to send and receive RF at the same time and frequency, effectively doubling capacity. If true, it would make small cells into something more like relay stations - definitely one to watch.
Lime Microsystems seem to be challenging the major RF front end chipset vendors with a solution for LTE that allows great flexibility through a programmable interface. I'm hearing this has been used for several unusual and unique applications, making it easier to stretch the boundaries of what 4G can provide.
Do you remember Mission Impossible (this tape will self-destruct in 5 seconds....)? Boeing have developed a super secure phone that will do exactly that if stolen. Of course, that's fine as long as you didn't share all your information with Facebook and Google first ;)
Next year's event will again be in Barcelona, but shifted back a week to the first week of March (2-5) because Chinese New Year falls mid-February.
Apologies for omitting many details and announcements from the show due to pressure of space. I'll be following up on many of the topics covered in this report and adding more detail in upcoming weeks. Let me know what I should be including or expanding on - just comment below (can be anonymously).