Fixed wireless access (FWA) continues to make inroads into markets big and small – as it has done for many years. Maybe that is why recent announcements of fixed point-to-multipoint (PMP) high-capacity mmWave 5G trials and the deployment of pioneering national wireless-to the-x (WTTx) rural broadband networks took our FWA experts on a trip down memory lane – some a bit further down than others.
Before we follow them, we should point out that the members of the Real Wireless team are as generous in their view of FWA as the next person – assuming that next person is a highly trained, engineering-savvy telecoms consultant who has worked in wireless telecoms at many levels.
But quite a few of them do have a bit more hands-on experience of FWA than most –and are only too aware that this is a technology as demanding as it can be rewarding. Perhaps that’s why, during a recent conversation on the topic, utopian visions of fully FWA-connected countries provoked a few looks back – to a time, in fact, when the sort of tools and processes Real Wireless now uses to assess wireless technology offerings were still a few years off.
But what does FWA in rural areas have in common with FWA in central London or parts of the US, both of which came up in the conversation among our experts? More than you might think. First of all, the two questions with any FWA deployment are: Can it be done? And will it make any money? And that applies to any rollout. You’d think, for example, that providing FWA for city centre businesses would be less of a problem than planning for several hundred consumers in a rural village where even one hardly utilised site could throw your calculations out.
Not quite. One of our younger experts, Oliver Bosshard, couldn’t resist pointing out that even densely populated London could be problematic. Oliver spoke from experience: he was involved in just such an FWA project before joining Real Wireless. Even then, the planning tools were fairly primitive compared to what Real Wireless now offers. On the other hand, there was likely to be a reasonable return on investment as many of the customers were businesses and willing to pay a premium for the service.
However, ensuring that you can earn that ROI isn’t easy when you have to identify potential and real demand, find rooftops able to offer line-of-sight connections and then work out whether these sites can serve enough people to pay back your outlay. Not to mention assessing whether a relay can be used (and what type of relay). And that’s without considering permits, buildings obscuring other buildings, the cost of buying expensive new map data (high resolution 3D data for cities) and re-running all the information, uncertainty about addresses that haven’t signed up but that might after the sites are built and, of course, changing skylines. You don’t get many cranes crowding out your line of sight in rural areas, after all.
And of course, rural consumers probably don’t need as much convincing when it comes to decent broadband. Oliver pointed out that his potential London enterprise customers needed to be educated in wireless concepts such as redundant connectivity, load-balancing or fail-over switches. Some needed help getting landlord permits for antennas in the roof. There was to be some sharing, and thus there had to be conversations around ‘clear-channel’ versus ‘low contention’, comparisons with ADSL, queries about private IPv4 and IPv6 addresses etc.
On the other hand at least the London project got off the ground, as have many others, rural and urban, in recent years. That said, John Okas, who has been around a bit longer than Oliver, offered examples of his own from an almost antediluvian period (or at least that’s how it seemed to Oliver) of how tricky FWA can be. One was a 1997 business plan based on 3.5 GHz destined for FWA by Ofcom. That got precisely nowhere. Then there was the would-be American prospect for FWA as an adjunct to fibre who allowed John into a meeting long enough to say he had stopped using wireless links due to rain fade, snow outages and general unreliability. Then the meeting ended.
But Oliver was not to be outdone. He threw in a few more challenges: the cost of rent, lease and staff – and just about everything else – in central London. Finding skilled technical staff able to work long hours. Having to get even more of them to ensure that business customers don’t suffer outages or, if they do, have very short downtime.
That’s not all. Using 5 GHz licence-exempt or light-licensed spectrum resulted in higher overheads to deal with increased noise floors, interference and other problems. Using 28 GHz was less of a challenge in some ways but increased the operational and planning demands. Both meant paying out for yet more highly skilled engineers. All that and hoping hardware gets delivered, spectrum is released, enough capacity can be delivered and funding is available (did I mention that some of this happened during the credit crunch?).
And don’t forget those damn cranes.
Now it’s perfectly reasonable to argue that rain, snow, cranes and credit are not necessarily going to affect a rural broadband FWA network. However, even if, as John said to Oliver “I was doing FWA business plans whilst you were still in nappies”, both would admit that, even while the planning has got more sophisticated, so the technology and business model have become more demanding. And, as John pointed out to Oliver, where these business plans often go wrong is that unobtrusive cell in the spreadsheet labelled ‘Market Share’ by year. Optimistic growth plans forget that most customers already have a service – so you need to make ‘conquest sales’ and persuade them to leave their current supplier. It’s easy to say: “Let’s assume 5% by year three.” In practice it can be a nightmare to achieve (let alone to achieve profitably).
Clearly FWA is, and always has been, something of a learning experience. It’s also still a popular answer to certain connectivity needs. What we know today is that FWA started as WLL (wireless local loop), then became BWA (broadband wireless access) and has, more recently, even been labelled as 5G in the mmWave bands. In the US it is often called FWB. But it is still, typically, a PMP architecture – and has its very own economic and practical challenges, of which line-of-sight limitations are probably the most talked-about but by no means the only ones – as this short overview has probably made clear!
So we’re still learning – which is, after all, part of our job, and the reason why we are constantly developing or refining our tools and processes to deal with the challenges that FWA poses. None of us at Real Wireless assumes that the FWA rulebook is written. Many hurdles have been overcome since 1997 but FWA continues to come up with new ones. Overall it’s true that FWA can be a great architecture and technology in the right place. And there are many challenges. But, summed up as simply as possible, making it work, practically and financially, is still the main one.