As I participated in a MEF webinar on 3 Sep on “The new mobility: WiFi6 and 5G” together with the CEO of the Mobile Ecosystem Forum and the CEO of WBA (the Wireless Broadband Association), I thought it would be appropriate to summarize some of my perspectives on the topic here (not all covered in the debate though).
6 is obviously greater than 5, so does this mean that WiFi6 will take over and make 5G redundant? I have previously commented on the hypes around 5G a few times (e.g. here and here), and also on the perspective that every odd “G” is not really needed. When this is said, 5G is definitively happening – but the question of who will eventually benefit - and what will happen to the ecosystem - remain open. Who will come out as most successful in the consumer market and in the enterprise market could also differ. There are various considerations.
Technology (or industry) wars:
Some might position this topic as a technology war – but I am not sure it is. Anyway, “technology wars” are not new to me. It is something I have seen ever since I, as a young engineer, was part of developing 2G and 3G in the late 80s / early 90s. Since then, as part of the mobile (cellular) industry, I have seen the industry grow from practically zero to 8.8 billion connections worldwide today (including cellular IoT), with 5.1 billion unique subscribers – and I have seen 2G, 3G, 4G and 5G launches across the world since then. After 20 years of development, WiFi is today also rather ubiquitous in homes and offices globally – and I use it every day. Also a great success !
Some earlier examples of technology wars I can think of include 1) the 2G and 3G technology wars e.g. between FDMA, TDMA and CDMA, a “religious” debate about which technology was “best” (something I was clearly involved in), 2) the mobile versus satellite debate around the time when LEO satellites were introduced – and 3) a similar WiFi versus cellular debate at the time when WiFi started to creep into phones. I was not specifically engaged in this last one personally, but it resembles the questions raised at this stage – and Mobile World Live has previously also commented on it. Their conclusion is that WiF6 and 5G are complementary and will work together. I believe yes and no. It depends.
More significant than the technology itself is the business side of things, however. From the mobile (cellular) industry side, I have seen increased competition every year, between mobile operators, between telco vendors, between mobile operators and new entrants, how every new “G” opens up for new ecosystem players, how new entrants position themselves in the standards development process – and even how standards bodies and industry associations compete on wireless standards – not unlike the question potentially asked in this case. For some fun reading on the changing ecosystem, check out also the futuristic article written by an old friend of mine, Bill Best of Azenby, here. In any case, it is normally the business models, convenience and the commercial aspects that will decide on a technology’s success!
Except possibly in the minds of occasional researchers, technology wars are not really about technology. Most “technology wars” in the past have really been disguised commercial battles over industry dominance and technology leadership (like the 2G / 3G technology wars between different European telco vendors and their backing Administrations) – and in many cases these days they are actually also regional / geo-political battles - be it between USA and China, between (American) OTTs and the global mobile operator industry - or otherwise. For more on geo-politics, see also one of my previous articles here.
I have not been directly involved in any specification work on 5G nor on WiF6 (so I could have missed something crucial). However, to the topic at hand, as is to be expected, WiFi6 is clearly better than WiFi5. This includes greater speeds (up to 10 Gbps?), increased throughput (4x more per user in dense environments?), increased network efficiency (4x?) – and increased battery life (7x?). This is of a similar nature to the promises of 5G, with quoted numbers like 10x lower latency, 100x faster than 4G, 100x more capacity than 4G, 1000x more devices, etc. To compare, we could easily claim that 10 Gbps with WiFi6 is 10x the 1 Gbps 5G expects. However, it is not only a question of how broad the band is – but also how “long” the broadband is. In other words, it is also about the capabilities of the backhaul and core networks – which are not provided by hotspot providers and have nothing to with WiFi. So is 6 better than 5? Not necessarily !
There are several key differences between WiFi and cellular – simply because they have been designed for different purposes. They are different from a network perspective, from a user perspective – and also from a coverage perspective.
WiFi is an access network - and will always be an access network. 5G is a mobile system technology, i.e. it has a core network with mobility features and much more – in addition to a new RAN. WiFi is positioned around products, i.e. routers or Access Points – and is focused very much on product sales to consumers or businesses. In this context, backwards compatibility with existing devices in the field is a key element. 5G, on the other hand, although it is also much focused around new smartphones for consumers, in the larger picture it is about mobile network infrastructure and mobile services. Another difference is that cellular predominantly is designed for wide-area and outdoor use (although it works perfectly indoors), with a range of tens of kilometres – while WiFi is predominantly designed for indoor use (hundreds of meters). Thus, they don’t really compare – but that does not mean they don’t necessarily compete.
From a technical perspective, as an engineer by training, I believe in Shannon’s law and the laws of physics, i.e. you can optimize any type of technology, but there is a limit to how much you can squeeze into the physical resources you have, be it time, frequency or space – and energy - and ultimately you will reach the same limit.
The consumer market and the enterprise market:
As I have commented in some of my earlier articles, 5G is a new “G” that can do everything 4G does, just better – and, in addition, it promises to revolutionize industries and societies through digitalization. Many (me included) have previously also commented that operators need to capture the B2B segment with 5G, i.e. the digitalization of industries – simply because the consumer smartphone segment will not provide sufficient return on investment. Quite simply, I don’t expect consumers to be willing to pay more for “even more broadband” – but rather less, as we have seen year after year. Operators on their side, however, will have to build another network, quite different this time – and with a lot of backhaul, in addition to maintaining several previous “G”s for a long time … and to make it worse, operators are not really good at it, i.e. dealing with vertical industries, SLAs and enterprise solutions for the B2B segment. Most operators are mainly focused on pushing smartphone to consumers (which represents the volume of business).
While 5G and smartphones most likely will provide “even more broadband” and ubiquitous wide-area coverage for consumers with great success, including also WiFi6 functionality, I would assume that WiFi6 could have very good prospects in the enterprise market – with wireless connectivity and solutions with high bandwidth and low latency (similar to the 5G URLLC use cases) in the industry segment, due to convenience, ease of deployment and given availability of good industry solutions. This remains to be seen, but at least, operators need to get their act together on industrial use cases. WiFi6 could thus provide competition for 5G from a use case perspective in some use cases.
We could say that the mobile industry is a slow industry, like an elephant. It takes time to move (e.g. once every 10 years) – but when it first moves, it is really difficult to stop. This goes for the SDOs, the telco vendors as well as operators, however, also academia and regulators have a stake in this game, wanting to cash in on research funds or on spectrum assignments. To date, this has provided great success for all parties involved. Let us just hope that the elephant does not run full speed into the wall. Will the elephant make it – or will the lighter and more agile mouse beat it in the end?
The topic of WiFi6 versus 5G becomes even more interesting when we factor in OpenRoaming, an initiative by WBA to create an ecosystem between WiFi hotspots, establishing something similar to automatic cellular roaming. We have all tried in the past (me included – even though I have been working for mobile operators for three decades) to use WiFi and avoid using the cellular service whenever possible, be it at home, in the field – or when roaming abroad – simply because it has been expensive. On the other hand, we have all also experienced the hassle of logging on to potentially insecure WiFi hotspots in hotels, coffee shops or airports – and having to pay a 10-dollar fee per day (or similar).
Things have changed, however. The commercial barriers associated with cellular are more or less gone. Mobile operators have bundled domestic mobile data as part of the monthly subscriptions, often with practically unlimited offers – and mobile roaming charges are now also often “free” (i.e. bundled with your domestic use) – at least in the EU and EEA – and recently also in CEE and in Central Africa. While WiFi hotspots traditionally have been a hassle with logon and dedicated usage payment, cellular data should really be seamless – and at reasonable cost these days – and you don’t need another service provider!
On the hotspot side, things have changed as well. The usual 10 dollars a day is less and less frequently seen – as hotspots have more and more given up the consumer charging. The hassle, however, remains – but this is something OpenRoaming is trying to fix – as it promises to provide seamless hotspot roaming.
Commercially, however, I have more open questions. As far as I can understand, OpenRoaming is in many ways aiming to replicate the cellular roaming regime, with “Identity providers” (i.e. service providers with billing functionality – like a mobile operator), “Ecosystem brokers” (e.g. wholesale interconnect carriers or roaming hubs) – and “WiFi network providers” (i.e. hotspot owners) with access networks. A main difference, however, is likely that, while cellular roaming primarily is focused on international roaming between mobile operators, the initial focus for OpenRoaming would probably be national - covering a wide number of domestic hotspots (although probably not exclusively). Anyone can take any role in this ecosystem, which is most likely not regulated, including also mobile operators – and we can expect that many different types of companies will position themselves in the area.
International roaming in cellular networks have been a major success factor for the global reach and scale of the mobile (cellular) industry today. Technically it just works – and commercially the consumers don’t see much trouble (except for some historical cases of bill shock – which are mostly gone today). What consumers don’t see, however, is the complexity and time it takes to establish commercial roaming agreements between all the mobile operators – which to a large degree have needed updating for every new service to be included in the roaming. This is a main reason why roaming hubs exist. With or without roaming hubs, however, the cellular roaming regime desperately needs to be significantly simplified (a revolution) to be viable in the long-term! Going forward, in my view, “roaming” could be something totally different in a longer-term 5G setting – maybe some other form of “service continuity”?
With OpenRoaming I would expect that “Identity providers” will take the role as a national service providers connecting a wide number of local hotspots to their ”networks” using standard technical interfaces and a standard commercial agreement. On an international level, however, I would expect that, in most cases, agreements between “Identity providers” will be needed for global ubiquity – and it will be necessary for OpenRoaming not to replicate the complexity of cellular roaming! It should be noted that nothing prevents an “Identity provider” to connect hotspots across the globe directly (maybe Google? – or large operator groups?).
The widening ecosystem:
While OpenRoaming has been established by WBA, mainly driven by non-mobile operators, we have also seen several large mobile operator groups like Orange, Deutsche Telekom, AT&T etc embracing OpenRoaming. These operators provide WiFi hotspots already – and they are also service providers (identity providers) and also “ecosystem brokers”.
For an operator there are several options for making use of WiFi. Mobile operators already today integrate WiFi into their cellular offerings – and for 5G, 3GPP has defined “non-3GPP access” as a 5G access network – as a means to offload traffic onto WiFi when convenient. I would expect that international WiFi roaming thus also easily can be part of cellular roaming – as long as roaming agreements allow for it. Domestically, operators may or may not already integrate their own WiFi hotspots – and / or they could use OpenRoaming to connect others, acting as “Identity providers”.
With the opening of the OpenRoaming ecosystem for anyone to take on roles, anyone can be a service provider (“Identity provider”), including existing mobile operators. From a user perspective, in the larger context, there might be a question of how many service providers you want to have a relationship with. If you already have a mobile operator, do you need another service provider? This applies whether the mobile operator provides WiFi access or not – in particular if cellular access is available and perceived “free”.
If you are not a current licensed mobile operator, then OpenRoaming may be an opportunity to expand your business into wireless service provision - without any consideration about 5G – and I would imagine many companies will have a go at this, including any kind of OTT player like e.g. Google, large hotspot providers and many others. It will obviously also create new business for device and equipment providers and others.
Security is a concern in wireless networks – and in a range of areas. This includes authentication of users, encryption of communication and more – and, in a larger context, denial of service, privacy breaches, fraud, espionage and more could be the result. Not all of this has been relevant for WiFi in the past, but mobile operators have dealt with all these issues for years – and one of the strengths of mobile networks is secure and trusted authentication through the use of the SIM. Further, without getting into details, the 5G standard has a number of security improvements over 4G. However, opening the ecosystem further than 4G might also increase vulnerabilities. For WiFi and the open internet, the ecosystem is already open.
In the case of WiFi6 and/or OpenRoaming, it is important to ensure secure authentication. With OpenRoaming this seems to be the case, provided the authentication provided by Identity providers is strong enough. For a mobile operator using WiFi6 in a 5G network, this will be the case automatically with the SIM. Mobile networks further encrypt communication, at least over the RAN, and with extended mechanisms is 5G. OpenRoaming has also been designed with encryption capabilities. A challenge in both cases will be lawful intercept, however.
While the battlefield in the field is likely to be ease of use, commercial relationships and costs for the consumers or enterprses, the framework for operation, and indirectly the capabilities of the technology, is also critical – and a question on the table is availability of spectrum. WiFi, as an unlicensed technology, operates today in the 2.4 and 5 GHz bands. However, in USA, the FCC earlier this year opened the whole 6 GHz band (5.925 – 7.125 MHz) for unlicensed use – and following this, Google requested permission to do trials across the US in the band aimed at developing a Spectrum Access System (SAS) to manage spectrum sharing automatically.
The situation in other regions of the world is, however, different. China has opened the whole band for licensed use (i.e. 5G) – and Europe has opened only the bottom 500 MHz for unlicensed use. The rest will be for 5G. A consideration in Europe is that the upper part of the band today is widely used for microwave backhaul – thus cannot be opened for unlicensed use – and the 6 GHz band is the lowest band available for microwave, thus the most affordable. All in all, we can expect the 6 GHz band to be a battleground for licensed or unlicensed use going forward – and up to WRC23 this will become intensified and will be another geo-political battle. Next week’s Light Reading webinar on the 6GHz band for 5G is an evidence of this.
So will WiFi6 be the end of 5G – or will they work together?
I have no doubt that both technologies will be widespread. 5G will clearly happen – and it IS happening already. It has a lot of momentum with operators, telco vendors, governments and regulators. The open question remains, however, if any operator will make money from it. So far, what we have seen from 5G is mostly on consumer broadband with smartphones and Fixed Wireless Access for fixed line replacement – which can be sold to a global consumer market.
The enterprise (B2B) market promise with digitalization of industries and society, which operators need to have for return on the 5G investment, could, however, be more of a battlefield going forward – and here the operators may not only struggle commercially with private 5G networks – but also possibly with WiFi6 entering the market.
Comparing the two technologies directly makes little sense. Technically, WiFi6 and 5G are complementary – and from a technology perspective, both have advantages and limitations. Operators will embrace WiFi6 as an upgrade to the WiFi they already see and use, as an access network in 5G or in other ways. Other players will focus purely on WiFi6 – and at the end of the day, what will count for consumers or enterprises are convenience, ease of use, and cost.
What happens to the 6 GHz band at global level remains to be seen. Whether it goes to licensed or unlicensed use will enhance the prospects for which ever technology it goes to – but otherwise it does not change much. So far, it seems only the US (and the UK?) have opened it fully for unlicensed use. In addition to technical capabilities, the outcome will have impact on economy of scale.
A very interesting aspect for the future evolution of the wireless ecosystem is OpenRoaming and how it plays out regarding interworking and ubiquity of service. There will probably be new players entering the wireless service provision arena – and mobile operators are also likely to embrace OpenRoaming and to integrate it into their offerings somehow. The 5G standard allows for “non-3GPP access” as a 5G access network – and operators are likely to use this somehow.
For most users, national ubiquity of service will suffice. However, as a seasoned international traveller, the main challenge for me in the past has been logging on to WiFi hotspots in hotels or airports – so for me, OpenRomaing would need to cover also the international roaming aspect – or I would simply use my existing cellular roaming service.
Interesting times for the wireless industry development going forward ….