The recently deployed WiFi 6 network at São Paulo/Guarulhos International Airport in Brazil has been touted as the first airport-wide installation of this sixth generation of the technology.
Airports generally have been a proving ground for WiFi 6 in public arenas, and an environment in which WiFi and 5G will be duking it out to support private networks of huge complexity. For WiFi 6, deployments such as São Paulo can help to establish whether the added benefits really do wow passengers and overcome limitations of networks based on earlier versions, such as contention delay for access, often stuttering performance and concerns over security. But the value of airports as testbeds has been compromised during the Covid-19 pandemic by a dramatic drop in passenger numbers so that networks have hardly been stressed as they sometimes were before lockdowns and travelling restrictions began.
The Sao Paolo airport network was installed by Boingo Wireless, which is in the vanguard of WiFi 6 networks. This is the biggest airport WiFi 6 network so far, and would have served around 50m passengers in a normal year. It does however come over a year after Boingo’s first such WiFi 6 deployment at John Wayne airport serving the Los Angeles area in the USA.
Despite serving fewer passengers at present, the company regards airports as an important yardstick for WiFi 6 to establish its ability to sustain the quality of service consumers have come to expect, able to stream and download high definition video content as well as movies. Such capabilities have been intermittent for earlier generations of WiFi and are seen as essential to match the promises if not realities of 5G.
That said, LTE often, in practice, exceeds the performance of WiFi in public venues so long as a decent signal can be obtained, partly because bandwidth either over the wireless network or the backhaul broadband pipe is often unable to meet demand, especially at peak times. In some cases, far from offloading, the service is being onloaded to the cellular network.
The question now is whether WiFi 6 alone fixes those WiFi growing and scaling pains resulting from rapidly increasing demand in such venues. The answer is that it will buy time because it will be several years at least before 5G’s millimeter wave bands become widely available for consumer services at such venues.
For now, WiFi 6 is good enough to meet requirements both of passengers and airports’ administrative applications, which often sharing the same campus network. However, the large areas covered and increasingly stringent demands for bit-rate and latency of some emerging applications around automation and remote control will demand performance superior to WiFi 6 capabilities, over separate spectrum. This is where private LTE or 5G networks will come in, operating perhaps over unlicensed or shared spectrum.
WiFi 6 on its own has been designed to extract more performance from the existing two WiFi bands at 2.4 GHz (2400 to 2495 MHz) and 5 GHz (5170 to 5835 MHz). The latter band initially provided relief from the congested 2.4 GHz band but has now itself filled up, although that is less of an issue for most airport spaces.
However, there can be an issue with the use of wider 160 MHz channels under both WiFi 5 and WiFi 6 because this reduces channel diversity given the relatively limited total spectral bandwidth available. Channel diversity is crucial for optimal performance because, over congested networks especially, multipath fading can diminish the available bit-rate over a given channel at any time. By hopping onto another channel, as is supported by the WiFi standard, consistent quality of service can be maintained, but only if there are enough channels available.
Improvements such as use of multiuser OFMDA (MU-OFDMA) enable WiFi 6 to sustain headline data rates about 35% higher than WiFi 5, but more importantly average data throughput three or four times higher through more efficient traffic management. Similarly, the ability of the network to sustain high resolution video streaming for multiple users will depend not just on the WiFi’s aggregate throughput but also on capacity of the broadband backhaul.
Nevertheless, as WiFi 6 is incorporated into ever more smartphones, laptops and IoT devices, as well as the WiFi access points (APs) and routers themselves, either operators or owners of public arenas like airports, business campuses and shopping malls will be able to deliver higher speeds, greater capacity and lower latency to more simultaneous users as well as devices. Airports are proving the ideal surrogate for such environments by combining a range of use cases within a single campus, if not quite under one roof.
WiFi 6 will be unable in its own to meet all these use cases and will coexist increasingly with private cellular networks for the more demanding enterprise applications under the banner of industry 4.0. The next iteration, WiFi 6E, will help shore up the existing WiFi user base, even if it makes limited inroads into the industrial wireless domain.
WiFi 6E improves performance and robustness by operating in the newly allocated 6 GHz band from 5.925 to 7.125 GHz. Apart from increased raw capacity this brings several other benefits, such as much greater channel diversity for frequency hopping. WiFi 6E brings 14 additional 80 MHz channels plus 7 160 MHz channels, which then combines the extra capacity with the resiliency of channel diversity for adjacent access points.
This gain will often be amplified by avoidance of contention with legacy devices that will be unable to operate in the new 6 GHz channels. The sharing of the network with legacy devices is referred to as the ‘bad apple’ problem, taking the analogy of one rotten fruit contaminating all the others. In the case of WiFi over 2.4 GHz, and to some extent 5 GHz bands, legacy devices tend to hog the channel when transmitting or sending data because of their lower data rates. They take longer to transmit a given segment of data.
In environments such as stadiums and airports that can be heavily populated, this combination of channel diversity and avoidance of the bad apple problem alone will improve the user experience enormously at peak times, even without the added capacity.