The cellular industry sometimes uses a mantra “Build it and they will come”. It certainly felt that way at the Moscone Centre which was undergoing some serious expansion last week - the noise of piledriving next door was hard to miss. This was the first MWC Americas since the GSMA had taken over organising it from CTIA, and there were many other changes to the format and content apart from the major building work.
The guiding hand and experience of the GSMA conference organisers was clearly visible compared to my previous trip to CTIA a few years ago but I expect it will take another couple of years before the event can truly establish itself as the leading mobile event in North America. Attendance of 21,000 was lower than the 30,000 hoped for. I viewed this very much as US centric with relatively little on LATAM despite real-time translation into Spanish and Portuguese.
The venue itself is split across three different buildings and several floors, with multiple entrances. Although much smaller than the main MWC event in Barcelona, it was quite hard to navigate with limited signage. But as Monica Paolini commented, there wasn’t much point in getting too familiar with the layout since the event is moving to Los Angeles next year.
Stands are considerably downsized compared with MWC Barcelona with even the major vendors taking a more modest approach. There was a substantial area for smaller companies and startups which was difficult to navigate but did provide at least a small desktop space and focal point for each. Companies choosing only a meeting room might be less visible but were still quite active and productive. Resellers primarily interested in accessories and smartphone covers were given their own area which kept the main trade floor halls less busy.
Keynotes
GSMA chairman Mats Granryd gave a potted history of the past 10 years in mobile, emphasising the rapid take-up of LTE worldwide which now has 2.3 Billion users. Their latest GSMA Mobile Economy Report for North America forecasts that the cellular industry will grow to 4.7% of the US GDP by 2020, today employing 2. 5 million people and contributing $110 billion in public funding.
This is very much a North American event, with patriotic keynotes emphasising that the region had led in 4G and expecting to retain leadership for 5G (whatever that turns out to be). CTIA CEO Meredith Attwell Baker asserted that “the US has the best mobile experience”, that 5G will create 3 million jobs and add $500 Billion to the US economy. She did note that 5G will require small cells on streetlamps and encouraged the FCC to give clear direction to States (regarding planning/zoning laws) to facilitate the forecast $275 Billion investment.
Carlos Slim Domit spoke about the success of America Moviles across Latin America, reminding us of the poor state of the network before they took over and the huge investments they've made. The contrast in ARPU between Canada ($48), USA ($39) and LATAM ($7) remains stark. He expects data traffic to continue to grow dramatically.
FCC Chair Amjit Pau spoke enthusiastically, addressing first the sterling effort to restore outages after the recent hurricanes. Apparently only 5% of cellsites were affected in the Houston area compared to 25% during Hurricane Sandy. He attributed that partly to the wider deployment of fibre backhaul vs copper, with better preparation and foresight by all operators and the FCC itself. Many in the audience were listening for clarification about CBRS, such as when we might expect approval for commercial launch (hopefully before end 2017) and what the rules around Priority Access Licences might be (perhaps more than 6 years maximum). There were encouraging signs but no new formal announcements. He is aware that rural coverage can be poor or non-existent and pointed to the Mobility Fund Phase 2 (worth $4.5 Billion) available to bring 4G to non-covered areas.
CBRS
CBRS was for me the hottest topic of the show, and I’ll be publishing a separate report of the CBRS seminar held at the show. Some key points were:
- FCC expecting to approve the system for commercial operation around end 2017. Initial services will use the General Access tier, with the rules for Priority Access Licences likely to be revised. The maximum 6 year tenure for area licences may be extended to 10 years to help justify larger investments.
- Outdoor service likely to be first application, particularly Private LTE and using USB dongles.
- Strong support from operators (e.g. Verizon) and cable companies. Charter Communications are actively running trials with 8 different vendors and are likely to launch next year.
- CBRS Alliance has delegated a certification testing process for small cells to CTIA, similar to that for Wi-Fi Alliance.
- Federated Wireless, probably the leading Spectrum Allocation Server provider, announced a round of $42 million investment, including from Arris, Charter Communications and American Tower. They will deploy environmental sensors nationwide by mid 2018. Service will be available immediately after FCC approval in inland areas, with coastal areas becoming live during 1H 2018 where the remote environmental sensors have been installed.
- Many new vendors have developed CBRS compatible small cells - just look at the CBRS Alliance member list for clues.
5G
Naturally there was a lot of talk about 5G and several demos, including with Virtual Reality. The most attractive use case seemed to be for fixed wireless access, effectively allowing cellular operators to compete with wireline/cable operators for residential broadband access, bypassing the last mile of copper. This is thought to be relevant for more densely populated areas, leaving sparsely populated regions to be served by 4G.
With so much 5G hype around, it remains to be seen how this will mature.
LTE network densification
There are many ways to increase network capacity and several different options were on show. North American operators are insisting that all new outdoor equipment is at least 4x4 MIMO, and Kathrein were showing off their 10-port pole-mounted canister antenna that incorporates the usual 1.8-2.6GHz cellular with both 3.5GHz and 5.8GHz bands (for CBRS and LAA respectively). Even if these new frequencies aren’t widely deployed yet, new urban sites are being equipped to support it.
Sprint and Ericsson demonstrated Massive MIMO (with up to 64 transmit and receive channels) giving significant increase to each sector even with today’s 2x2 MIMO smartphones. Samsung were also showing a extensive range of kit with similar capabilities. This approach emphasises that operators will continue to make the most of their existing cellsites and licenced spectrum where possible. The advantage is that it uses the existing handsets and spectrum, but does not achieve faster Gigabit data rates (although 40-50Mbps per handset is quite feasible). The demo screen below shows 340 Mbps total throughput from a single 20MHz LTE carrier.
Where much faster data rates per handset are required, LAA (combining 10 or 20MHz of licenced bands with up to 60MHz of 5.8GHz unlicenced) is commercially deployed in San Francisco and other cities across the US. Qualcomm showed me a live demo downloading large files (1Gbyte) in just a few seconds using one of many commercially available smartphones. Theoretical data rates of 1Gbps are possible and I’d expect several 100Mbps quite feasible in the real world. Fully capable smartphones are likely to be more pricey than average.
What struck me walking around the streets were the large and somewhat ugly antenna installations in the city today. These already have a lot of sectors, see below left. The picture below middle was taken at LAX airport, where the short term car park has been used to mount many antennas, yet the signal from some operators inside the building was quite poor. A better indoor solution is surely needed. There seems to be growing interest in strand mounted units which can be hung from street wiring, although I doubt that would be the best choice in regions likely to be affected by storms. Several DAS vendors were illustrating disguised street side formats such as the matching "BigBelly" garbage bin format by JMA.
Indoor Repeaters
It’s clear that there remain many customers poorly served by networks.
The ability to buy and installer your own equipment to solve indoor coverage or quality issues remains a quick-fix choice for smaller to medium sized buildings.
SureCall are one of several companies legally offering 1 Watt repeater units that can be self-installed (or by suitably trained technicians) at prices below $2000. They appeared to be attracting quite a lot of interest at their stand, suggesting that there remains no shortage of poorly served buildings.
Nextivity Cel-Fi have seen a lot of interest in their latest QUATRA product, which is a smaller off-air DAS solution with four radio nodes serving small to medium sized properties. It can also be driven directly by a small cell (or multiple small cells for multiple operators).
Mixing DAS and Small Cell
Driving a DAS system directly from a Small Cell is quite viable, especially where the Small Cell is already approved and integrated into an operator’s network. Verizon have approved SpiderCloud for that purpose, and Corning (who have just acquired them) may develop a more direct and efficient digital interface. Additionally, SpiderCloud can be used to provide the 5.8GHz LAA capability alongside. Art King of SpiderCloud told me that they see great potential for CBRS and are actively involved in trials. Bill Cune of Corning sees good opportunities for joint solutions but re-affirmed that both products will be available for sale independently to interwork with any other vendor equipment.
Other DAS vendors continue to enjoy ongoing upgrades to existing installations even if fewer new buildings are being equipped. Several different strategies are being adopted. Chinese vendor Comba have both small cell and DAS capabilities but focus on the latter in this region. JMA remain independent. SOLID offers its own built-in “signal source” but has not yet got this approved by any operator.
Zinwave have taken a different approach to serving CBRS and/or LAA with a DAS system, developing a separate DAS radio head for the 3-6GHz band. This could be deployed alongside their existing 0-3GHz unit. Although I’ve seen a few DAS vendors consider expanding support to 3.5GHz or even 4GHz, I’ve not heard of this approach elsewhere.
Other in-building solutions
Distributed radio systems included the Ericsson Radio Dot on show in its various formats (see below) as we reported last week, with the option to install in under venue seating. Chinese vendors ZTE (below right) and Baicells were also exhibiting a wide range of products with the latter having made inroads into over a hundred of the smallest independent network operators across the country.
ASOCS launched their Cyrus solution which is an LTE baseband unit with digital output capable of driving DAS or other external radio heads. It can also host Mobile Edge Computing applications and is aimed at the Enterprise. Their differentiator is being built on standard Intel hardware and a VMWare virtual platform with a dedicated hardware accelerator board for the RF processing. As with any small cell vendor, certification and integration into operator’s networks will be their main challenge to gain market acceptance.
MulteFire Alliance observed that the same use cases for Private LTE that are driving CBRS also exist elsewhere in the world and are a good match for MulteFire. They’ve just released V1.1 of their specification which shifts the focus a little to incorporate some IoT capabilities. Their 35 members now include device vendors such as Samsung, Sony and Sierra Wireless. The technology will first require both LAA (downlink) and eLAA (uplink) formats at 5.8GHz in both small cells and handsets. Their timescale is soft launch before end 2018, initially with USB dongles.
Open Network Foundation
I had an extensive briefing from the ONF who are creating a reference design for 4G and 5G networks based on virtualisation. One distinguishing factor is that contributors can provide either fully open source code or resell binary images. The open source code is free but may be less fully featured and comply with older standard releases. You get what you pay for.
Demos included a live network slicing implementation using Accelleran’s CBRS small cell – allowing clear resource allocation of an enterprise cellular radio between in-house and external roaming users.
Summary
A worthwhile show with potential to become the leading North American cellular event. I don’t yet see this incorporating LATAM, which in any case has quite different drivers and demands. Features such as CBRS, the significance of cable operators to the cellular market and other regional differences justify focussing this as a USA/Canadian event.
The sprinkle of GSMA gold-dust has helped improve and remodel it, but it needs a bit more work to achieve its full potential.
Small cells for Enterprise, especially CBRS and for directly driving DAS, were significant opportunities for new and existing vendors alike.
Next year's MWC Americas will be held in Los Angeles between 14-16 September 2018.
I'll leave you with a view of the double-decker Oakland Bay Bridge below rather than the more famous Golden Gate one.
Comments
RSS feed for comments to this post