Despite FWA hype, mobility is the urgent 5G goal, even in mmWave

AT&T demonstrated surprising hesitancy last week (in public at least) about its business model for its initial 5G deployment. Marachel Knight, SVP of wireless network architecture, told the Mobile Future Forward conference that the operator would launch initial 5G services by the end of next year, but has not yet decided which services to support.

The company is vying with Verizon to get 5G into the field, though their approaches are very different. Verizon has been clear that its first deployments will be for fixed wireless access (FWA) only, with mobility to follow later; and it is using its own pre-5G specifications, rather than waiting for official 3GPP standards to be frozen (though it insists its network will be easily tweaked later to conform fully to those standards). It has tested FWA in 11 markets already but has not announced details of its commercial plans, such as pricing or data rates.

AT&T, by contrast, led a group of companies which persuaded 3GPP to accelerate the work on one strand of the 5G New Radio (NR) specifications in order to get them completed by the end of 2017. That will enable the operator to launch in the same timeframe as Verizon, but with standardized equipment, and no need to make later adjustments in order to be 3GPP-compliant. The initial fast-tracked standards are for 5G NR Non-Standalone (which requires an LTE host network, and so is only deployable by current MNOs).

However, AT&T’s ideas on services seem less developed than its rival’s. When asked, at the event (organized by Chetan Sharma Consulting), whether the initial network would support fixed or mobile operations, Knight said: “We haven’t defined that use case yet.”

In the early days of 4G, some operators launched fixed services first, because that enabled them to get to grips with the new network at an early stage, but without initially having to worry about smartphone availability or mobile hand-off and roaming. The same is likely to be seen in 5G, but the big two US operators have additional motivation to support FWA –  as a way to replace ageing DSL links where fibre is not cost-effectives (AT&T is already doing this with LTE); and more importantly, a means of extending fixed-line and quad play services outside their wireline territories.

However, this is far more interesting to Verizon than AT&T, because the former’s wireline territory is far smaller, confined to the densely populated and wealthy, but fairly small and saturated, north-eastern region of the USA. For most operators though – and even for AT&T outside a few select Verizon markets – FWA is a niche use case. Mobile services, delivered to consumer and enterprise smartphones, tablets and cars, are the low hanging fruit, because broadband and media usage is shifting from fixed to mobile, and because the MNOs already have these customers signed up. For most, mobile broadband will remain the core of their business case for years to come, with 5G only being optimized for other businesses – notably in the Internet of Things, enhanced wholesale services and critical communications – in a second phase.

The other very early 5G movers, such as NTT Docomo and Softbank in Japan, and the two major South Korean operators, intend to support smartphone services from day one, even if they use pre-standard equipment, because they will be showcasing 5G at Olympic events in 2018 and 2020.

Though it will be challenging to support full mobility at this early stage, this is where the money will lie, so AT&T may be thinking that sticking to FWA risks a damp squib in terms of market impact. After all, T-Mobile USA says it plans to deploy mobile 5G as early as mid-2019, and has scoffed at the FWA plans. Outspoken CEO John Legere said, when he announced his own firm’s 5G roadmap, that the big two were “using 5G to either distract from how badly they’re losing today or to give their shareholders some hope they can compete with Big Cable. Their ambitious vision for fixed 5G to replace home internet will never provide mobile 5G coverage. It makes no sense.”

Knight agreed that mobile 5G is “where the industry ultimately needs to get to”, and this may be easier for AT&T than for Verizon, since 5G NR Non-Standalone leaves mobility management and wide area coverage to the 4G anchor. Knight did not provide a date when AT&T intends to move to the standalone standard, which includes a full core network with full user and control plane capability for 5G.

Not that AT&T is backing away from its FWA plans – it just may include mobile services too, which would help it outshine Verizon in the publicity stakes (after Verizon took the crown in LTE, becoming the US’s first operator, and the world’s second after Telia, to deploy commercial 4G back in late 2010). AT&T said last week that it would expand its 5G FWA trials to three more cities – Waco, Texas; Kalamazoo, Michigan; and South Bend, Indiana – by the end of this year, in addition to existing tests in Austin, Texas.

Like Verizon’s, these FWA trials are using millimeter wave spectrum, in 28 GHz and 39 GHz, which supports high data rates, but over short distances. For wide area mobile services, sub-6 GHz spectrum is likely to be preferred, at least in the early days while mmWave radios remain immature.

The challenge of making mmWave spectrum, with its temptingly large capacity, as usable for cellular networks as sub-6 GHz bands, is a centrepiece of the large vendors’ R&D programs. Ericsson and Nokia both made announcements in this area, as part of larger 5G-related launches, last week. Ericsson is a believer when it comes to FWA as a ‘pioneer’ business case for 5G. Dave Allen, a distinguished engineer with the vendor, said in a keynote at the Hot Interconnects event earlier this month: “It will be easier to plop a pole in a neighborhood than connect homes via fiber.”

Ericsson, which remains a mobile network vendor at heart despite forays into convergence, says the huge capacity boost – which 5G radio designs, and new spectrum such as mmWave, will enable – will see wireless traffic overtake that on wireline networks by 2027. But this will mean some complex technology becoming workable and mainstream, especially in mmWave bands, where transmitters and receivers will need to use Massive MIMO antennas with beamforming to achieve peak performance – they will compensate for the signal loss of about 40dB in 39 GHz compared to 3 GHz. These advanced technologies are needed to enhance range, allow signals to bounce off buildings and mitigate other limitations of high frequency spectrum, in order to improve performance and avoid interference – which, at mmWave, can come from “radio passing through my body, flocks of birds and trucks driving by…there’s 10m times higher loss than on wired nets,” as Allen put it.

Hence the focus on Massive MIMO which, as reported by LightReading, he described as “a form of spatial multiplexing, assigning multiple low speed signals to different antennas in an array, limited by the smallest number of antennas on either end.” Combined with beamforming, spectrum reuse can be maximized, though it will be hard to get these technologies into handsets. “Handsets will calculate math transforms to stay focused on a tower using no moving parts, just by changing the phase and manipulating signals to reposition themselves instantly,” Allen explained.

Meanwhile, base stations will use CoMP (coordinated multipoint) – part of the 4G standards but enhanced in 5G – to relay beamforming jobs among sites.

Elsewhere, Ericsson was putting its money (or its antennas) where its mouth is, releasing its first Massive MIMO 5G product for FDD spectrum. Like most 5G NR launches from all the major vendors, this is designed as a 4G-to-5G bridge product, with suppliers well aware that, more than in any previous technology generation, operators will deploy 5G tactically where it can enhance their business case, while continuing to enhance 4G as much as possible for many years.

Interworking, parallel roll-out on common infrastructure, common virtualized core networks, even spectrum sharing between the two technologies – all these will be central to the plans of most MNOs, and the vendors need to support that. So Ericsson’s new AIR 3246 for Massive MIMO, which will be commercially available in the second quarter of next year, supports both LTE and 5G NR, and is the Swedish company’s first 5G NR offering for paired spectrum.

Most Massive MIMO developments, for 4G and 5G, have initially been in high frequency, single-band TDD (unpaired) spectrum, high frequency single-band TDD because this requires smaller elements, making it easier to build large arrays. Massive MIMO beamforming is more efficient in TD-LTE because the unpaired spectrum allows easier dynamic feedback between device and network. But even in FDD, the technology can treble network capacity and boost user throughput by five times.

Ericsson has three TDD radios capable of supporting 5G and Massive MIMO in its collection. But it has lagged behind Huawei, ZTE and Samsung in unveiling FDD Massive MIMO. Operators like China Unicom, Softbank and Telkomsel of Indonesia have trialled Massive MIMO in FDD-LTE, with the Chinese vendors, in recent months, while AT&T has invested in 3D beamforming start-up Blue Danube, whose 96-element BeamCraft500 active antenna has been tested in dual-band FDD spectrum.

This is not all about mmWave bands. Some trials are taking place in mid-band spectrum, which will be the workhorse of 5G in the early years, especially as global allocations and band plans for mmWave, let alone national regulatory environments and auctions, are yet to be clarified. Ericsson’s first announced trial of its new FDD Massive MIMO solution is in mid-band spectrum with T-Mobile USA, on three LTE sites in Baltimore, Maryland. “T-Mobile’s racing forward at breakneck pace with Ericsson’s next-gen tech that advances LTE today and paves the way for 5G tomorrow,” said TMO’s CTO Neville Ray in Ericsson’s press release.

So Massive MIMO is becoming a classic bridge technology, allowing operators to choose to enhance the capacity and performance of their LTE network, or move to 5G (or both) when it suits their business case. Joe Madden, founder of mobile networks research firm Mobile Experts, wrote in a blog post in February: “Instead of waiting for 5G, mobile operators are upgrading thousands of TD-LTE base stations with Massive MIMO this year. It’s an exciting development which, as it gains momentum, may be able to leap into the FDD market as well. Blue Danube’s recent FDD field trial results using simple modules are very promising, and represent a big step toward cost effective FDD and 5G solutions.”

Ericsson is eyeing the same transitional market – which could last for a decade. “The new radio will enable operators to enhance 4G capacity for their subscribers today and be ready for 5G tomorrow, using the same hardware,” said Fredrik Jejdling, head of business area networks at Ericsson. “We also complement the products with a set of network services, simplifying the journey to 5G for our customers.”

Stefan Pongratz, senior director at infrastructure analyst firm Dell’Oro Group, believes that 10m LTE macro radios will have been installed by the end of 2021, despite most operators still focusing heavily on 4G expansion and optimization. He said: “Just as carrier aggregation has been key to adding needed capacity to mobile broadband networks, Massive MIMO has the potential to be the primary capacity enabler in the next upgrade phase, providing a smooth transition towards 5G.”

Ericsson’s main rivals are also focused on Massive MIMO and beamforming, as well as 4G/5G coexistence. Huawei recently announced that it had achieved peak cell throughput of more than 20Gbps in the 4.9 GHz band in a Chinese test using 5G NR, Massive MIMO, polar code and other technologies (see Wireless Watch September 4 2017).

A day after Ericsson’s launch, Nokia followed suit, expanding its small cell (see separate item) and macrocell portfolios with increased capabilities in MIMO, beamforming and carrier aggregation. It was more 4G-focused in its announcements, though all the products are designed to provide stepping stones to 5G too.

The Finnish vendor extended its AirScale Remote Radio Head family to boost peak performance and cell capacity while reducing space requirements on towers. New dual- band and single-band FDD-LTE and TD-LTE radio units support carrier aggregation (including unlicensed and shared spectrum), 4×4 MIMO and 8×4 beamforming (supporting a two-layer transmitter MIMO array of 16 RF channels and 32 digital channels). These radios are also designed to support a wider variety of bands, and higher output power, plus simplifying network deployment.

Nokia also announced an agreement with Qualcomm to conduct interoperability testing and over-the-air field trials based on 5G NR in mmWave bands. Qualcomm has just announced its prototype system based on initial specs emerging from 3GPP for Release 15 (the first 5G release), which will be finalized by year end. The prototype system operates in bands above 24 GHz (Qualcomm already offers one for sub-6 GHz spectrum) and features the chip vendor’s mmWave RF front end design in a smartphone form factor with MIMO, adaptive analog beamforming and beamtracking/beamsteering techniques. It supports 800 MHz of bandwidth and 5G NR technologies such as LDPC channel coding for data channels. The Nokia collaboration will be based on the Finnish firm’s 5G FIRST and AirScale solutions.

Samsung and KDDI beat mobile handover record with at 192km/hour:

Samsung hopes that 5G will give it an opportunity to get back into the mobile infrastructure game in earnest, leveraging its strong track record in small cells and virtualization to move into the mainstream.

Last week, it announced the results of a series of 5G tests with KDDI of Japan, which took place on a race track and demonstrated mobile hand-off at a speed of 192 kilometers per hour, the fastest ever in mmWave spectrum.

Conducted at Everland Speedway in South Korea, the test measured and evaluated metrics including handover interruption time, uplink and downlink throughput stability, and latency stability.

Woojune Kim, head of next generation strategy in Samsung’s Networks business, said: “It is becoming increasingly important that we accelerate our focus on 5G’s ability to meet a growing number of performance metrics. Until now, peak bandwidth has been the common refrain, and certainly a big component of the future of 5G. However, the test we conducted with KDDI will help us build a more diverse portfolio of future 5G use cases.”

Akira Matsunaga, senior director of mobile network technical development at KDDI, said: “The trial showcased stable performance under high speed mobility conditions which will dramatically increase the service experience of users in vehicles.”

The partners are making high mobility a centrepiece of their trials. In February, they demonstrated 5G handover in an urban environment at speeds of up to 60km per hour on public roads.

Samsung said in a statement: “5G’s ability to deliver on its promise is directly tied to its ability to meet the unique performance needs of a wide number of increasingly connected industries, from smart city IoT to connected cars and more.”

UK’s EE doesn’t wait for 5G to offer fixed wireless:

It is not just 5G which can be used to fill gaps in wireline broadband networks. BT’s mobile subsidiary, EE, has launched an LTE-Advanced wireless router which it says will bring downlink speeds of up to 90Mbps to UK users over its 4G network.

Of course, LTE has been used for fixed wireless already, but usually in very rural or underserved areas. In the early days of 4G, some operators deployed it – rather like Verizon now – in order to start work on the new network early, before smartphones were readily available and without having to address issues of roaming and full mobility.

But early claims that 4G would be a viable contender to fibre were quickly discounted as fibre providers pushed their optical links closer and closer to the premises, while boosting speeds and quality of service.

The same may well prove true of 5G, but cellular does have a place in filling the gaps in the fibre network and reaching rural or remote areas. In EE’s case, it has announced its LTE Category 7 4GEE Home Router, which can connect up to 32 devices at once. It will be targeted at users in areas of poor fixed-line access.

Max Taylor, head of marketing at EE, said that the solution was “a great alternative to fixed-line broadband” and was “simple to set up and use straight out of the box, so it’s ideal for customers with slower fixed line speed or people who regularly move house and need instant WiFi connectivity”.

It seems likely there will be a 5G equivalent once EE starts to implement next generation technology, enabling BT to address customers even where its fibre does not reach, and to satisfy government demands for universal access (the government only actually mandates 10Mbps speeds).

The router is available for free on an 18-month contract with prices ranging from £25 a month for 10GB of data, to £100 a month, for 200GB. The device can also be bought upfront for £129.99 with 10GB of data pre-installed.

EE’s LTE network covers 83% of the UK’s landmass, or 99% of the country’s population. It is aiming for 95% geographic coverage by 2020 and is the only UK MNO to release this statistic. It has been upgrading its network to LTE-Advanced, including refarming more 1.8 GHz spectrum to boost capacity. LTE-A is now available in more than 150 towns and cities.