knowledge centre

Arqiva and Samsung demonstrate 5G FWA while Openreach launches an FTTP demand consultation

Rupert Wood Research Director

"mmWave may be credible enough to get fixed operators to by-pass time-consuming demand assessment exercises for FTTP."

This comment contrasts the rate at which 5G fixed–wireless is developing with FTTP in countries where FTTP has not been widely rolled out, and argues that the reality of competition should – and doubtless will – supersede slow demand assessment exercises in ultrafast broadband roll-out.

mmWave fixed–wireless access arrives in Europe

On 27 July 2017, UK towerco Arqiva and Samsung demonstrated a pre-5G fixed–wireless system in central London, which they claim to be the first of its kind in Europe. Several versions have already been trialled in the USA.

The demonstration shows a single 1.1Gbps downlink over 230 metres using 300MHz of 28GHz spectrum, for which Arqiva already has licences (in part through its acquisition of Luminet's 2×112MHz the previous week). Massive MIMO and beamforming mean that that capacity can to some extent (unknown to us) be spatially multiplexed, so this figure should not be taken as the total shared capacity of the cell. The total capacity at the cell could in theory be as high as 10Gbps with 800MHz of spectrum, and can serve a maximum of 64 connections. While in the demo the cell antennas are mounted on a rooftop, the proposed solution has a unit mounted on a lamppost with remote baseband. Arqiva already has concessions with many municipal authorities for access to street furniture. The signal is delivered to a small indoor CPE unit next to a window without recourse to external antennas (see Figure 1).

Figure 1: CPE equipment for mmWave FWA [Source: Analysys Mason, 2017]

We visited during a biblical downpour, but there was no signal loss. However, the angle of incidence of the signal to the window does make the signal deteriorate, and the demo had the angle at close to the optimal 90 degrees. Building penetration loss can be turned to the service provider's advantage in the system, as the signal will bounce off properties in urban environments, extending NLoS range.

Arqiva and Samsung claim the system is cheaper than pulling fibre into a property, and far faster to roll out. In dense urban areas like central London, we would assume a typical cost of GBP300 for a final fibre connection (fibre and ONT installation). In the mmWave system, the comparison would be with the shared cost of the cell, plus the CPE antenna, plus any installation costs. In addition, the mmWave system will have fibre backhaul, and the density of fibre required will probably be greater than that in an FTTC deployment, so it will require new fibre in many cases.

Broadband investment decisions have come to a crunch-point in the UK and other FTTC countries

Openreach has indicated that improving FTTP economics make passing 10 million premises (about 40% of UK) a possibility, and is gauging interest from broadband service providers (SPs), enterprise plays, MNOs and others, particularly in the emerging 5G landscape.1 The consultation document states that Openreach's understanding is that the only major operator that is planning to replace widespread FTTC with FTTP is Proximus in Belgium. In fact, several overlays are already underway (see Figure 2).

Figure 2: FTTC-to-FTTP overlays underway [Source: Analysys Mason, 2017]

Operator Location Planned FTTP premises passed already passed by FTTC
Bell Canada Toronto and Montreal At least 2 million
AT&T Various cities 14 million out of about 30 million passed
TIM Various cities Up to 5 million
Proximus Urban Belgium 2.5 million (50% of Belgium)
Skipti Reykjavik 75 000 (50% of Iceland)

The main driver of these FTTC-to-FTTP overlays so far is not demand for gigabit connectivity, but competition from other infrastructure-based players. Openreach's first 2 million premises tranche was we think a response to new FTTP build, particularly in business districts, by other players, most notably Virgin Media.

Openreach faces political pressure to expand FTTP, but the UK is not unique in this respect. In addition, Openreach faces regulatory and commercial pressures in three areas.

  • Openreach will be subject to a price cap on wholesale FTTC that will reduce the basic tariff by GBP3 per month.
  • New duct and pole access regulation (DPA2) should make it easier and cheaper for other FTTP and fixed wireless providers to lay their own fibre in Openreach physical infrastructure, even if the condition of that infrastructure means that it is unlikely to be as easy to roll out fibre as in Spain and Portugal. Investment in emerging fibre players like CityFibre is ramping up, and DPA2 improves their investment case.
  • Hutchison 3G UK (Three UK) has indicated that it wants to roll out a new fixed wireless network covering 40% of the UK.

Arqiva's activity adds to this pressure, and what is more, Arqiva is for sale.

It is understandable that Openreach wants a consultation for a major capex project (GBP400 per premises passed looks a reasonable benchmark for FTTC-to-FTTP overlay). But any network investment model has to take into account not only the risk that demand will not materialise soon enough, but also the risk of doing nothing. Those latter risks are building up, not just for Openreach, but for a whole host of other fixed incumbents that have not yet made the commitment to FTTP, such as Deutsche Telekom's German business and, ironically, NBN.

We have our doubts about mmWave 5G FWA.

  • It is not cheap, and even if capex is lower than FTTP, opex will be higher.
  • It may be difficult to market (Arqiva will wholesale only) because realistic end-user speeds for individual properties will be difficult to predict (mid-hundreds of megabits per second looks likely, we think).
  • It is also difficult to predict what proportion of properties would need external antennas (some undoubtedly would).
  • It will be a massive challenge for a new player to scale up to attract sufficient tenants.

But mmWave FWA has many advantages: it can be rolled out fast, it can deliver ultrafast bitrates, and it paves the way for 5G mobile. Some operators are investing now; in acquiring Straightpath, Verizon has just spent about USD22 per premises in the USA for national mmWave coverage.

Coupled with other emerging competitive threats, it may be credible enough to get fixed operators to by-pass time-consuming demand assessment exercises.

1 See