Fixed wireless access (FWA) has been the belle of the ball in the first phase of 5G, largely because it is the only mainstream use case that is enabled by the early 5G platforms, that provides some operators with a brand new revenue stream. While the enterprise and IoT services will need to wait for the 5G core and future standards, and the mobile broadband use cases are more of the same, FWA offers selected operators a chance to provide fixed/mobile convergence and multiplay services for the first time.
The long term business case looks marginal for most operators, however. Three UK is one of a handful of mobile-only providers which are launching FWA-first, to leverage plentiful amounts of spectrum to offer generous data and content bundles which work in fixed or mobile mode, in the home or outside. But few operators have as much mid-band spectrum as Three and, once their networks start to get loaded up with mobile traffic, will need to move up to millimeter wave if they are serious about supporting significant numbers of fixed users without damaging their mobile QoS.
The risk for FWA is that it will always have to compete with fiber which, once it gets built out to the premise or home, will always be a performance step ahead of wireless. There are windows of opportunity for FWA in countries where FTTH has not got far yet (like the UK), or where it is very expensive. But in a few years’ time, it is likely that FWA will be thrown back to the same difficult models that made fixed WiMAX tough – to undercut fiber and be forever stuck in price wars; or to target areas unserved by fiber, which brings high operating cost and, often, a low income user base.
There may be more enterprise cases for FWA which can use the improved speeds and cost of ownership of 5G, especially in premises that are hard to equip with fiber. But very few operators, despite the early popularity of FWA, are putting forward a fully standalone business case for fixed wireless as Verizon has – and the early ratings of Verizon’s mmWave build-out have been poor, with very few households in its target areas able to access the service. The company, whose business case rests on the fact that it has a very small fiber territory and so can expand to new territories with 5G, has just announced a new iteration of its service which may address the early teething troubles, including the much-needed self-install CPE.
In markets where there isn’t much fiber or users are very cost-sensitive, the MNOs’ opportunity will be weakened by the emergence of more open solutions. Some of these are coming from the outdoor WiFi community, using souped-up implementations of the 802.11 radio; others are coming from the Facebook-initiated Telecom Infra Project (TIP). TIP has something of a split personality. On one hand, it has efforts which are heavily driven by the established cellular operators, geared to driving open interfaces and hardware to support new economics for 4G and 5G. On the other, it has platforms like Terragraph, which are being used for FWA in unlicensed spectrum, providing the kind of low cost platforms that can empower challengers to the MNOs.
All these sometimes conflicting trends are, however, reigniting interest in FWA, and a buzz that has not been seen since the start of WiMAX 15 years ago. Whether the emerging architectures, 5G or otherwise, go the way of WiMAX (and before it, the LMDS technologies and the Wireless Local Loop) and become solutions for niche, underserved markets remains to be seen, but there are many choices available for old or new operators interested in this space.
At the recent WISP Alliance show in the USA, there were several FWA launches:
- Nokia launched a FastMile 5G Gateway which integrates multiple 5G bands and WiFi 6, and also upgraded its FastMile 4G Gateway. It says it is trialling the 5G option with 30 operators and starting to deploy with Optus of Australia.
- Infinet Wireless unveiled its new Axion 28 portfolio, designed for multipoint operation in the 28 GHz band; and added a 70 GHz (E-Band) solution.
- Siklu, originally one of a group of start-ups which focused millimeter wave technology on small cell backhaul, launched a new EtherHaul 2-foot dual-band antenna designed to work with its ExtendMM feature on EtherHaul 8010FX, 2500FX and other E-Band models. ExtendMM is a product suite consisting of the new antenna, software for monitoring and provisioning and a built-in switch in the EtherHaul radios. This can be integrated with Siklu SmartHaul apps such as LBC for link planning and WiNDE for network design.
Meanwhile, Qualcomm recently announced that over 30 OEMs – including Nokia and Samsung – are using its Snapdragon X55 5G Modem-RF chipset for commercial CPE which will be deployed for FWA services from 2020. The firm announced a reference design for 5G FWA earlier this year, integrating the modem, RF transceiver, RF front end and antenna modules and targeting sub-6 GHz or mmWave bands.
Qualcomm president Cristiano Amon said: “The widespread adoption of our modem-to-antenna solution translates into enhanced fixed broadband services and additional opportunities to utilize 5G network infrastructure for broad coverage in urban, suburban and rural environments. Due to the development ease of our integrated system and industry movement toward self-installed, plug and-play CPE devices, we expect OEMs will be able to support fixed broadband deployments beginning in 2020.”
Last month, Qualcomm announced an extended range implementation featuring the QTM527 mmWave antenna module for the Snapdragon X55 5G Modem-RF System. It seems likely that Verizon will be a customer for Qualcomm’s new high-power CPE design for Power Class 1 – Verizon has said it has been waiting for a CPE with higher output power to improve performance of its 5G FWA offering.
At an investor conference last month, Ronan Dunne, group CEO of Verizon Consumer, said the company would be ready for the mass market with 5G FWA when it had fulfilled two requirements. One is fully standards-based New Radio equipment rather than the telco’s pre-standard platform (it is in customer trials of NR now). The other is the high-powered CPE.