While Wi-Fi vs. cellular arguments are as old as Wi-Fi itself (cellular’s got a lot more miles on it, having gone live in 1983 in the US, just as the first way-before-Wi-Fi WLANs were beginning to emerge), a comment earlier this year from a senior executive at a major cellular carrier that 5G would displace and perhaps even eventually replace Wi-Fi reminded me of former Soviet Premier Nikita Khrushchev’s infamous proclamation of 1956: “we will bury you” – referring here, of course, to America. Utterings on such a grand scale, thankfully, seldom foretell the future – and, in this case, have no fear: Wi-Fi will continue to flourish even as carrier-based services grow with 5G.
Still, senior executives presumably get paid to add value well beyond merely stirring the pot, so let’s explore the strategic landscape to see why this pronouncement will never come to pass.
First, will end-users abandon the unlicensed bands to use only the licensed services provisioned by the carriers? Nope, doesn’t seem likely. Even with the restrictions essential in the rules defining usage of the unlicensed bands, Wi-Fi has proven that the end-user provisioning of reliable, high-capacity mobile broadband services is easy and effective – and cheap, given that there’s no billions of dollars in spectrum licensing fees to amortize. And the recent unlicensed availability of 6 GHz. spectrum adds even more capacity, again with no cost to use the airwaves.
So, then – why not deploy 5G in the unlicensed bands? Such could indeed be done, as we’ve seen with 4G (LTE-U/LAA), but would represent, in operation, serious interference to nearby Wi-Fi (and other unlicensed) systems. There are billions of deployed Wi-Fi devices humming right along today, and nothing helps to grow your customer base better than irritating those potential customers, right? Sure, Wi-Fi includes lots of features like DFS and other mechanisms to mitigate co-channel interference, and 5G could add its own, but I’ve always had mixed feelings about operating for-profit services in the unlicensed bands regardless, and very negative feelings about operating conflicting services in the same band/channel simultaneously. End-user customers can of course still decide what infrastructure will be deployed in their facilities – Wi-Fi , 5G, whatever – applying the prudence necessary to avoid capacity-sapping interference. And prudence here means deploying unlicensed 5G only if absolutely necessary, and avoiding interference to current infrastructure at all costs.
Next, how long will it take to deploy 5G services with the availability, capacity, and reliability sufficient to replace Wi-Fi? The carriers might have a good long-term pitch here: I’ve long believed that the best option for the carriers to grow going forward is for them to become managed services providers (MSPs), and thus ISPs, IT Cloud-services firms, and etc. Carriers are well-positioned to market these capabilities, and acquisitions could allow them to enter this high-value-add business without too much trouble. I believe that many organizations will continue to outsource their computation, data storage, and networking needs, a trend initiated during the 2008/9 recession and clearly enhanced by the current economic mess, and they’re already buying communications services from the carriers. But, again, how long might this take? Um, at least a decade, and likely much longer. After all, just deploying all those 5G microcells for the WWAN as required by current carrier business plans will take a very long time (and cost billions) all by itself, let alone learning the MSP business.
And, finally (for the moment), what is 5G? Sure, there’s a new airlink, 5G-NR. But, wonder of wonders, 5G at its core is simply the adoption of a wide variety of technologies already proven by Wi-Fi – smaller cells, wider channels, MIMO, more efficient modulation, improved management and operations, and etc. And, if we define 5G generically as an airlink capable, at least in theory, of throughput to any single user of 1 Gbps or greater, it’s easy to reach a perhaps-surprising conclusion: Wi-Fi 5 and 6 are, in fact, 5G! And before you scoff at that with “Wi-Fi ain’t cellular” on your lips, OK, granted: cellular is designed to provision stationary but most importantly vehicular (high-motion) communications across the wide area. Wi-Fi is better at low-motion communications across a more limited geography. But, other than that, Wi-Fi can indeed do everything that cellular can, and, again, at a lower cost because unlicensed spectrum adds no additional economic load. Wi-Fi, in short, doesn’t need to be replaced, and there’s no value for end-users in so doing.
Anything else, then? Security, perhaps? Yes, 5G has much-improved security features over 4G. But Wi-Fi security has similarly made much progress over the years, and both technologies can take advantage of L3/L4 security capabilities, so no advantage one way or the other here. OK, then; wouldn’t it be nice to have a single technology to optimize handoffs between the licensed and unlicensed bands? Hotspot 2.0/Passpoint may play a far larger role here over time, with the value of Wi-Fi across all applications – including telephony and streaming video – preserved.
So, then, really? 5G is going to obviate the need for Wi-Fi, and, ultimately, replace Wi-Fi altogether? Seriously? As I think I’ve demonstrated here, no, it won’t; what many, even some in the analyst community, have missed in the advent of 5G is that 5G isn’t intended to replace Wi-Fi at all: it is, rather, designed to replace wire, as in residential broadband connections, and, via BYOD, enterprise voice and out-of-building data communications as well. Carry your gigabit+ link – but most importantly your good old smartphone – with you wherever you may roam.
In reality, however, you’ll still need several of those 5G links, since all those mobile devices and all that IoT in your house still needs connectivity even when you’re away. So as cellular continues to replace traditional POTS (and some fixed broadband as well), and thus bind the user to the carrier a bit more tightly, we see the true business model for 5G.
If Wi-Fi didn’t have that huge installed base, well, then, sure; we’d only need 5G. But since the opposite is the case today, expect most residential communications, including IoT, to be based on Wi-Fi, and thus for routers, the platform for most residential Wi-Fi, which enhance security and reliability regardless, to maintain their place. And, no, I don’t see many enterprises as likely to drop their wired broadband links to the outside world for 5G, although 5G will over time play at least a small role here.
OK, two final points: the carriers today have a lot of control over the feature sets of handsets and other devices. Suppose they decided simply not to support Wi-Fi in hardware at some point? Yes, that could happen – but it won’t. Customers, we believe, would revolt, and handset manufacturers are now quite decoupled from the carriers with respect to sales. Nope; that won’t happen.
And might some (presumable novel) wireless technology really and truly eventually replace Wi-Fi? Sure; maybe. Those WLANs from 1983 used infrared links, after all, and they’re history today. But such a revolution is at present very hard to imagine – Wi-Fi has clearly kept pace with advances in radio and network technology, and we expect such to continue. The installed base is massive and continues to grow. And, most importantly, Wi-Fi is inexpensive, cost-effective, and free from the control of carriers or any other massive entity, apart from the FCC and other regulators who really have been facilitators of unlicensed servcies over the years. Bottom line: Wi-Fi gets the job done, and will continue to do so. Cellular replacing Wi-Fi, then? Ain’t gonna happen. No way; no how.