White Paper Provided by Dean Bubley, Founder & Director of Disruptive Analysis:
Richard Bennett [1] of the High Tech Forum has recently written a paper about Wi-Fi, on behalf of CTIA. It is mostly an analysis and criticism of the way in which recent versions of the Wi-Fi standard use 6 GHz spectrum, claiming that it is inefficient and ineffective in delivering performance benefits for consumers. He contrasts it with cellular (3GPP 4G / 5G) technologies.
He has also done some basic testing of his own, comparing different generations of consumer-grade Wi-Fi products in terms of data throughput. He claims this shows reduced improvements over time.
To be fair, Bennett has an accomplished history as one of the figures behind Ethernet technology and he’s demonstrated a firm grasp of the technical fundamentals of wireless/radio protocols. Moreover, he’s shown a willingness to do real-world testing (albeit without lab-style rigour).
However, some of his recent commentary and analysis on 6 GHz spectrum is superficial and selective, especially when applied to the real-world market for wireless products and services, and the needs of consumers. For example, enterprise users of Wi-Fi are hardly mentioned. He also appears to deny the benefits – or even the trade-offs – for democratizing wireless technology, against the more centrally-planned philosophy of licensed spectrum and cellular carriers.
Location-specific spectrum needs
Some of the paper’s comments have an ironic twist. Bennett makes various observations about the Wi-Fi industry’s use of spectrum that actually have direct equivalents in the cellular world too. With reference to the upper part of the 6 GHz band, Bennett says:
“Considering that Wi-Fi is mostly deployed in the home/office or in buildings and gathering places, the perceived need for any additional spectrum, if any, should be limited to small geographic locations”.
If you take the view that certain spectrum is only needed by Wi-Fi in certain locations, then it is easy to construct a very similar argument for 5G public mobile as well. The busiest urban cells have 500-1000x the usage density of quieter rural areas, yet often have the same spectrum allocations. In other words, “the perceived need for any additional spectrum, if any, should be limited to small geographic locations”.
Yet while the Wi-Fi industry is happy to explore a variety of options for very-low and low- power, as well as standard power where permitted by an Automatic Frequency Coordination system, much of the mobile industry seems implacably opposed to anything other than full-power exclusive licenses and wide-area (national) coverage.
High demand often translates to specific public venues, high-traffic urban areas, or industrial sites requiring private networks for specialized site-wide applications. Technologies such as Citizens Broadcast Radio Service intended to “localize” spectrum access face opposition from carriers wanting to cover larger areas – though these same carriers then cite high localized demand as an argument for more exclusive spectrum nationwide, rather than acknowledging it as a strong case for CBRS’s value in areas of high traffic density or specialized need.
Claims that vendors 'no longer see value in the 6 GHz band'
A rather confusing part of the report discusses some commercially-available Wi-Fi 7 routers, some of which split the 6 GHz band into two parts. It asserts that:
“If companies are dividing the band because of lack of international consensus and effectively giving up on the upper 6 GHz band, then there is no advantage to making lower 7 unlicensed, even if more spectrum mattered.”
This is a non-sequitur. It is fairly clear that many manufacturers would prefer to create a single product that can be sold to customers both in countries with the full 1200MHz released for unlicensed use, and also those with just the lower part of the band available at present. In addition, creating a single 1200MHz RF (radio frequency) front-end is challenging, especially for engineering down to suitable cost points for residential units. It doesn’t imply “giving up on 6 GHz.” The cellular industry frequently splits up 5G bands in similar fashion, such as the regional variations in devices supporting the 3.3-4.2GHz range.
Similarly, he notes (as has the author of this article) that a number of products have Wi-Fi 7 certification, but do not actually support 6 GHz at all. These are principally made for the Chinese market, where the band is not available – but then exported to the US and Europe, as they are inexpensive and thus suitable for service providers wanting a “7” brand for marketing purposes, especially for lower-tier customers. None of that is an indication that 6 GHz is not useful – just that early products are expensive, and also in some cases need a larger power supply for first-generation chipsets.
Fixed wireless access
Various parts of the paper reference the market for fixed wireless access, and its recent rapid growth, especially in the US. It asserts that assigning the 6 GHz band for Wi- Fi use, which is predominantly used indoors at low power levels, sits in contradiction to the needs and potential of FWA.
“What good is an ultra-high-capacity network inside the home that can’t connect to the Internet because wireless ISPs can’t get the spectrum they need to connect those homes to the global network?”
This is a strawman argument. Firstly, the majority of US homes with gigabit-grade connections rely on fiber or cable, not wireless or satellite access. FWA is still a roughly 10% niche, even if it has shown some notable growth. Continued FTTP (fiber to the premise) network build-outs are likely to continue to generate competition in future, especially in urban areas. T-Mobile and Verizon have both become much more acquisitive in the fiber space this year. There are no FWA-primary broadband markets around the world, and nothing suggests the US will develop one either.
Secondly, there is very little evidence to suggest that FWA deployments and uptake are significantly limited by spectrum availability, especially at a nationwide level. In fact, much of the boom in FWA has been a result of operators seeking to monetise their existing underutilized 5G spectrum in certain areas – it is an incremental and emergent revenue stream, not even envisaged in early 5G concepts or standards.
Furthermore, numerous FWA vendors and service providers actually support Wi-Fi 7 and 6 GHz for indoor connectivity, including using 6 GHz bands. For example, Verizon Business’ FWA gateway features tri-band Wi-Fi access. Qualcomm sells a chipset for FWA routers that specifically supports this feature, as well as allowing the use of 5G in many sub-6 GHz bands and also in the mmWave range. Recent technology advances have also shown that mmWave 5G (for instance at 28GHz) can have significantly higher range for FWA than initially believed, so US operators have additional options for expansion if needed.
In other words, FWA (whether delivered with 5G or other technologies) to the home, and Wi-Fi 7 inside the home make a good combination.
In fact, the unlicensed 6 GHz band actually also enables more options for FWA, if used at standard power, enabled by permissions from an AFC database. This type of offer from network providers could either use versions of IEEE / Wi-Fi type technologies, unlicensed variants of 5G cellular, or older proprietary FWA technologies if they comply with the rules.
At a recent 6G conference in Brooklyn, a T-Mobile attendee asked a question about whether future versions of 5G / 6G could be made simpler, and optimised for FWA deployment, including in unlicensed bands such as 6 GHz. They noted that unlicensed spectrum is often well-suited to the needs of FWA – something that has been demonstrated in many other successful commercial deployments.
6 GHz decision-making
Another false assertion in the report is the accusation that the Federal Communications Commission’s decision-making about assigning 6 GHz for unlicensed use was hasty and prompted by the pandemic’s impact.
“FCC assigned 1200 MHz of new unlicensed spectrum to Wi-Fi in the second quarter of 2020, the time of quarantines, an economy in collapse, and working from home. Naturally, conditions pushed the Commission to respond to the crisis in an extravagant but unsustainable way.”
This conflicts with the history of that process, which involved early discussions and advocacy starting in 2017, an FCC's Notice of Proposed Rulemaking in October 2018, and detailed technology studies in 2019, prior to Covid’s emergence.
It also doesn’t align with the later assignments of the band in various other countries, especially in the Americas and Asia, which had plenty of time to assess options and reach the same conclusions and decisions.
In fact, what was hasty was the response of the cellular industry, which until that point had defined 5G’s lower frequency range as “sub-6 GHz.” It belatedly started lobbying to change that to sub-7GHz, with access to the 6 GHz band, only after the unlicensed concept was well advanced. The CTIA took its position from 2019 onwards, GSMA from 2020, while 3GPP did not embrace the band until Release 17, with work also starting in 2020.
Another section in the paper discusses how modeling – specifically Monte Carlo simulations – inform policymaking and regulatory decisions on spectrum bands such as 6 GHz. While this is partly true, Bennett’s examples overlook a wide range of other inputs, notably differing timing and the role of advocacy groups, as well as political and geopolitical pressures. (His own DIY Wi-Fi testing might also have benefited from something like Monte Carlo analysis to cover a broader set of input scenarios).
In particular, the GSMA has been hugely active in global advocacy on 6 GHz licensing for IMT since 2020, largely in an attempt to dissuade other administrations from following the US path. Ironically its own models (or its consultants’) employ the same “continuous traffic” concept that Bennett derides in the paper, along with a spurious multiplier variable called an “activity factor." It also ignored the large proportion of cellular traffic that is consumed indoors, and the unsuitability of 6 GHz for outdoor-to-indoor coverage.
To coordinate or not?
In one of the later sections of the paper, Bennett describes a lack of coordination and the Carrier Sense Multiple Access protocol as “Wi-Fi’s original sin,” although the paper recognises that recent variants’ use of Orthogonal Frequency Division Multiple Access protocols and better scheduling go some way to address this. At the center of the debate about Wi-Fi and 5G is the relevance of coordinated use of wireless resources vs. uncoordinated.
In general, Wi-Fi requires at least three “non-overlapping” channels, because access points can be deployed by anyone, without the need for mutual coordination. So, for instance, a building with 5 tenants, or a street with closely-packed homes, may have many independent Wi-Fi networks in operation. That means that for Wi-Fi 7’s 160MHz and 320MHz channels, a substantial total amount of spectrum is needed in aggregate.
In the final section of the paper, Bennett comments that “the approach of doubling channel width in every generation while dipping into the mid-band spectrum pool isn’t working.”
What this glosses over is that the larger channel widths are optional and deployment- specific. What makes sense in a factory using 8K cameras on robot inspection units is likely to be different from a college lecture-hall with hundreds of laptops, or a home Wi-Fi mesh deployment linked to a 5G mmWave fixed-access router.
The uncoordinated Wi-Fi model contrasts with licensed-spectrum technologies, where the licensee (such as a mobile carrier) can manage radio resources with more oversight, and greater levels of frequency reuse. However, to offset this control, most regulatory agencies such as the FCC want competition between at least three “non-overlapping” rival network operators – so again aggregate spectrum requirements are multiplied.
Ultimately this comes down to a debate about the value of “democratized” wireless and permissionless use, compared to the value of structured but centralized wireless services such as those seen in cellular networks. There is no single right answer.
Wi-Fi in higher frequencies
The paper’s comments about Wi-Fi failing to exploit unlicensed 60GHz or (somewhat bizarrely) optical communications have some merit, but this idea faces similar problems as mmWave for 5G. These are under consideration for future Wi-Fi variants, but it is unclear if they can become widely used.
In particular 60GHz has poor propagation, and does not go through walls easily. This – or worse, optical communications – would imply using an access point in every room of a house – presumably backhauled by 5GHz or 6 GHz, except in a handful of properties flood- wired with ethernet or fiber. Notably, the word “wall” does not appear once in the entire document.
To be fair, it is not unreasonable to wonder if future versions of Wi-Fi could allow for better localized coexistence between neighbors. Something similar is being considered for the next version of CBRS's General Authorized Access as well. We may see future enhancements to AFC or CBRS’s Spectrum Access System platforms that could facilitate this. However, that is an opportunity for innovators, rather than a reason to argue against today’s networks.
Backward compatibility
Bennett makes another fair point about the impact of old devices on overall system performance, because of the continued use of legacy Wi-Fi variants alongside newer ones. But that is also true for other technologies. Backwards compatibility always poses a challenge of this type, whether that relates to old analog TVs and radios, IoT and M2M systems using 2G or 3G cellular, Bluetooth and a variety of other devices in the same bands such as microwave ovens or garage door-openers.
It is also reasonable to assume that the Wi-Fi industry (and connected devices manufacturers) would happily sell replacement upgraded devices, given an incentive to do so. The access point vendors would also be happy to reclaim more value from the 2.4GHz band.
So, it would be fair to suggest that regulatory authorities might consider how such a “switch-off” and device migration process might occur in future – but again, that is a very separate question to that addressed in this paper.
Conclusion
In summary, the white paper is interesting but ends up arguing against itself. If it was positioned as a way of suggesting ideas for future Wi-Fi 8 or 9, and also for regulators wanting to maximize the economic and social benefits of unlicensed spectrum, it would have some value. It could also provide some useful input as policymakers consider additional frequencies in the 7-8GHz band for future licensed or unlicensed use.
But the central theme of the report – which seems to have an unstated objective of “re- litigating” the upper part of the 6 GHz band’s unlicensed status – is very flawed.
The report’s comparison of different access point performance is essentially anecdotal, using a single home and a limited set of tests. Performance gains are measured in many ways – not just throughput. There is no analysis of latency, nor changing behaviour under load from many devices.
We are still early in the stage of commercializing Wi-Fi in 6 GHz, with just the first generation of Wi-Fi 7 products and chipsets shipping. Most projections of uptake seem positive, and in particular the potential for dense environments and high-end enterprise use-cases have significant opportunities.
There is actually a strong synergy of unlicensed 6 GHz with FWA deployments, as it gives new options for both wide-area and indoor spectrum combinations. Together with carrier aggregation for 5G, and multi-link operation for Wi-Fi, we can maximize future optionality for many different scenarios and market outcomes – which is exactly what policymakers should be striving for.
About Dean Bubley
Dean Bubley (@disruptivedean) is the Founder of Disruptive Analysis. He is one of the leading analysts covering 5G, 6G, Wi-Fi, telco business models & regulation, and the emergence of technologies such as quantum networking and AI.
Footnote [1]
It should be noted that Bennett and this article’s author have sometimes engaged combatively via social media about wireless and spectrum issues. This piece is not connected to that, and is not intended as a personal attack. ↩︎
