This year saw SCWS World (formerly Small Cell World Summit) attracted somewhat fewer numbers in this, its 8th year despite moving to a very plush new venue on the south bank of London. I’m told the mix of vendors vs customers (operators and venue owners) has switched to have more of the latter. Certainly there were lots of meetings and networking between many of the regular attendees. Overall there was something lacking in terms of spark. The program was somewhat better than last year but still needs improvement with more structured, more lively and better content.
Perhaps the lower attendance was due to a clash with TM Forum or perhaps it partly reflects the stage of lifecycle of the business. There were fewer non-European participants, suggesting it has become more of a regional event. A pre-conference day on Mobile Edge Computing introduced a topic that is still in its infancy, with some quite lengthy and academic content. A series of case studies of practical deployments presented by Julius Robson of the Small Cell Forum were both useful and informative.
Workshops during the first morning were well attended and generally thought very useful. I was involved in sessions addressing unlicensed spectrum including LAA and MulteFire. Shared spectrum use in various bands including 3.5GHz for the US was of interest. This is as much for the potential new business models it might open up than any technical advantage.
Some commented that there is far too much focus on peak data rates that these schemes might achieve but which are often unachievable in the real world. Reliable and dependable coverage, especially for voice, is considered far more important by the average consumer. Actual and consistent data rates are more relevant.
Alan Law, Chair of the Small Cell Forum, opened with an upbeat message that small cell equipment revenues are now over $1 billion, and forecast to be $6 billion by 2020. A visible example comes from Reliance Jio of India who have an ambitious goal of 150,000 enterprise small cells. Alan forecasts that 90% of cellular traffic will be carried over relatively low cost small cells within the next five years, completely transforming the shape and format of cellular service delivery.
The Small Cell Forum Awards (for which I'm one of the judges) was the highlight of the week, rightly celebrating the success of both those shortlisted and the winners. ip.access swept three awards (Nick Johnson below pictured making yet another Oscar's speech), Parallel Wireless and Vodafone each won two.
Certainly the larger Enterprise market segment remains by far the most bullish. Cisco was very upbeat about their go-to-market strategy, noting that their larger Enterprise Wi-Fi customers appear to be prepared to pay the additional cost of Small Cell add-on capability during major Wi-Fi refresh/upgrades. Their channel partners can easily scale to substantial numbers of deployments that would not be feasible through operator’s own design and installation teams alone.
Huawei are also having a lot of success with their Lampsite indoor solution, with Version 2 now commercially available. Of the three major RAN vendors, they were most visible at the event with Ericsson and Nokia were both present but not sponsoring or exhibiting. Ray Williamson shared his perspective on how to grow the market.
AT&T’s David Orloff spoke of their Small Enterprise Self-Install option, whereby businesses can buy and self-install up to three of their own 3G/4G multimode small cells. Order to live commercial service timescale averages 72 hours – radically faster than when their own staff and processes are involved. They use Nokia’s (formerly Alcatel-Lucent) 8892 box which could be remotely upgraded to dual 4G when appropriate. Volumes are still relatively small and I could not easily find the option mentioned on the AT&T website, but it’s one of the few operators openly offering a commercial option for business customers wanting to solve their coverage or quality issues. David noted that achieved performance, especially coverage, often considerable surpassed the conservative figures of 15K sq feet per unit quoted in their marketing literature. In one exceptional case, over 120K sq feet had been lit up by just two units – it’s highly dependent on the local environment and construction materials, so there are no guarantees it would work that well everywhere. This will compete with Nextivity's Cel-Fi DUO+ repeater just released for use with Verizon from next month which also has a stated coverage area of 15K sq feet. AT&T also use an existing Nextivity 3G/4G multi-mode repeater product.
One panel session I moderated looked at the needs and expectations of public venues such as shopping malls. Trevor Pereira of Intu Properties (who operate several major UK shopping malls) spoke of their significant investment in (Free) Wi-Fi and DAS. While some 80% of visitors have Wi-Fi switched on (and so can be anonymously tracked as they move around the site), only around 15% bother to login. He has no visibility of how many or how often customers use the DAS but suspects it’s almost universal. Ideally, he’d like a solution that provides some of that visibility of what and where users are doing but with the universal scope that multi-operator solutions provide.
There was some debate about how MEC (Mobile Edge Computing) could solve this in the future. Privacy and security aspects, perceived as a major benefit by end users of cellular service, were noted – you would never get 100% of visitors to sign up. A Vodafone panellist said they thought MEC would be more successful for residential than enterprise.
This has been one of the slowest market segments to evolve, with operators first making the most of their existing assets and sites. Advertising giant JCDecaux was just one organisation touting access to a large potential range of sites.
iWireless guided me through their extensive planning and evaluation process, graphically modelling the City of London streets for both cellular and Wi-Fi service. The ability to plan and predict deployments is definitely available with product available. The view below illustrates modelling of outdoor coverage in the City of London.
There were relatively few independent wireless backhaul vendors exhibiting, most notably CCS and Intracom still in play. CCS now have an integrated unit which hosts a third party small cell and fits within the de-minimis (i.e. does not require planning permission) of Cambridge local authority. Both CSS and Intracom have self-organising solutions that simplify and add resilience to urban deployments.
Unlicensed and Shared Spectrum
Qualcomm gave a very good presentation on the evolution through LTE-U, LAA and MulteFire. In answer to my question, we can expect LTE-U hardware to be software upgradeable to LAA. While it seems to be this would mean LTE-U would have a relatively short lifetime (it’s advantage being mostly about time to market), this will be up to network operators themselves to determine. Small cells will be available very quickly but mass market smartphone devices will take longer to appear commercially. A few niche applications and equipment might appear in the second half, with real-world trials perhaps early next year.
Most small cells today have Wi-Fi built-in. It seems that the larger products would reconfigure the 5GHz radio to deliver LAA instead of Wi-Fi - I can imagine it would be challenging (if not counter productive) to try to operate both LAA and Wi-Fi in the same frequency band. I can only assume that handsets would be capable of either, but never actively transmit using both LAA or Wi-Fi simultaneously
The lack of multi-operator support with standard small cells has perhaps been its biggest Achilles heel. A simple solution is to deploy multiple sets of small cells – a so- called salt-and-pepper method. A ratio of 4:1 Wi-Fi access points to small cells (which have a wider coverage area) make this quite feasible. Finding some neutral spectrum or using a single network’s spectrum are equally technically possible. That’s generally been unachievable due to commercial intransigence of operators, but I heard of examples where this might become reality in the near future.
ip.access now have their Viper neutral host gateway in live commercial operation with JT (formerly Jersey Telecom). It runs on any server (and could run in the Cloud), so is just another application in their datacentre and can scale from 50 to millions of cells, supporting 2G, 3G and 4G. With a population of 100,000 Jersey (an island just south of the UK), the network celebrated connecting its 1 millionth customer last year. This highlights the potential for innovative smaller networks to move quickly into niche and growing alternative commercial opportunities.
In the past, I’ve spoken with a few players aiming to become neutral host businesses, acting as the middlemen between the few network operators and the larger numbers of individual enterprises. Many are not yet public. There will be various opportunities in the value chain, and I definitely believe this is one space to watch.
Both Cisco and ip.access are keen to point out that each vertical market in the Enterprise segment has its own unique requirements and channels. Cisco's Mark Gallagher speculated that in the future we might see different neutral hosts emerging, each focussing on their own specific enterprise vertical – different ones for hospitality vs business offices for example.
Snippets and Announcements
I’ve not included all press releases below, but here are some highlights
- The Small Cell Forum published the first few foundation documents of their HetNet and SON Release, covering a wide scope of long term HetNet architecture. A further Release dealing with Virtualisation will come out towards the end of 2016. You can’t fault them on the number or volume of documents published, although I do wonder if this detracts from the more practical focus of getting Small Cells adopted.
- Radisys launched TotaleNodeB+, which adds 3GPP Release 13 features including LAA to their standard small cell software solution. Their software is incorporated into many of today’s successful Small Cell products including those from Airspan, Parallel Wireless, Baicells and many others.
- Arcadyan were showing their latest multi-mode 3G and LTE enterprise product, which shares a single baseband chipset for both radios (pictured below on the right). There remain relatively few truly multimode Enterprise small cells available, with many LTE only vendors preferring to avoid investing in 3G. Markets for LTE only include USA, China, India and Japan while Europe, LATAM and Africa will need 3G compatibility for some time.
- Airspan were showing off an indoor cellular relay (below). This self contained "accordion styled" box incorporates both a full receiver and transmitter, retransmitting 4G service on a different frequency within a building. Unlike a simple repeater, the parameters of the inbuilt eNodeB are fully integrated into the network and configurable both centrally and through SON. Its similar functionality to that shown in their larger outdoor products, shrunk into a single large book size unit. This would be especially relevant for areas that don't have adequate wireline broadband, such as in India.
- Parallel Wireless displayed their growing range of small cell products. Their technology design is incredibly power efficient, radically reducing the size of their 5W+5W unit including heatsink to probably the smallest, lightest and lowest power in the industry. Software configurable, these can be used for 3G or 4G. CTO Rajesh Mishra told me he saw no reason they couldn’t support other technologies in the future too, such as NB-IoT (Narrow Band LTE for Internet of Things).
- BaiCells, a very new Chinese vendor (who have already captured the largest share of China Mobile’s TD-LTE nanocells), again demonstrated their LAA capable unit and their tiny ElfCell. They’re keen to point out they have support for FDD mode too, supporting European and/or other regional frequencies on request.
I’d like to see a clear structure for the conference program next year that separates Indoor (mostly Enterprise) and Outdoor (Urban, Rural, First Responder), which attract different audiences on different days. Perhaps a third day might look at future technologies. The workshop format was educational and of high value.
The date and venue has been set for next year already. It’s the same hotel (O2 Intercontinental London) on 23-25 May, so does not overlap with TM Forum.
I'll leave you with the stunning view across the Thames looking north towards Canary Wharf, the financial centre of London.