Your browser is not supported. Please update it.

19 March 2020

WiFi 6E pitched for coexistence with 5G

Apart from being an essential upgrade for streaming Ultra HD video at acceptable QoS, WiFi 6E, the latest iteration formerly known as IEEE 802.11 ax, is pitched at levelling the playing field against 5G. In a sense, it means that we are in for another era of mixed competition and cooperation but with some moves towards that long promised complete harmonization within homogenized heterogenous services.

Yet harmony relies on mutual respect and it had looked like 5G with its broad range of spectrums represented an attempt to steal WiFi’s clothes by enabling more reliable indoor coverage. That is why WiFi 6E was all the more important for closing any gap 5G had opened up over total capacity and headline speed.

That said, 5G’s indoor performance would not be adequate for high performance streaming video within the home to smart TVs, tablets and laptops, where WiFi, strengthened by this supposedly sixth generation technology, will reign supreme. 5G however may enable good enough indoor coverage for smartphone users such that they no longer resort to WiFi when indoors at some public venues, as they have been more or less forced to with 4G. That is assuming the constraints of restrictive data plans are finally confined to the dustbin of history.

At any rate, the WiFi community is now pushing WiFi 6E harder and urging regulators to open up the 6GHz band for unlicensed operation. This, according to the Wireless Broadband Alliance (WBA), the body dedicated to WiFi interoperability, will provide more capacity than all the other WiFi bands put together and support throughput equivalent to 5G.

Several regulators in North America and Europe are now considering release of 6GHz spectrum for such unlicensed use, leading wireless SoC (System on Chip) leaders to ramp up their activities and predict 500 million WiFi 6E compatible laptops and mobile devices will be available inside three years by the end of 2022. Broadcom has singled out streaming video as the key driver for WiFi 6E, followed at the opposite end of the bandwidth spectrum by the Internet of Things (IoT).

The WBA has been conducting trials in San Jose, California, using WiFi 6E-capable mobile platforms and laptops enabled by Broadcom and Intel silicon, to assess how close the technology comes to its promised bit rates and robustness. The top headline speed of WiFi 6E is 9.6 Gbps, up from 3.5 Gbps for WiFi 5, which was known as 802.11ac for most of its lifetime before the simplified generational naming scheme was introduced by the WiFi Alliance in 2018.

These are both theoretical speeds attainable only in perfect lab conditions that are never obtained in practice, but they set the bar for real performance and – in the case of WiFi 6E – the objective was to enable sufficiently high bit rates simultaneously to many more devices, while also shaving latency down to the low levels achieved by 5G at the cell level.

It was designed not just for homes and businesses but also public venues like hotels, airports and train stations, where performance and availability in the past have been notoriously unreliable. Video streaming is often impossible and sometimes banned in such situations.

Forthcoming tests over the next few months will evaluate performance in homes with challenging RF conditions resulting from thick walls or metal obstructions, as well as underground rail networks where the aim is to extend the signal through tunnels rather than just within stations. During the first trial phase, speeds of 2 Gbps were achieved, as well as two-millisecond latency connections, both comparable to 5G.

A little clarification is needed here because WiFi 6E is what WiFi 6 should have been and already seems to compromise the simplified naming scheme for standards where the number simply denotes the generation.

In fact, WiFi 6 at first only brought a nominal improvement over WiFi 5 in terms of capacity, continuing to use just the dual 2.4GHz and 5GHz radio bands. WiFi 6E, which probably should have been called WiFi 7 but came too soon after the basic WiFi 6 to be considered a generational step change, then added operation in the 6 GHz band as well, boosting potential capacity and performance significantly. This would have been added with WiFi 6, but at the time regulatory agencies would not allow that because the spectrum had been reserved for licensed purposes that could make money.

However, in October 2018, the US FCC first mooted offering the 6 GHz spectrum for WiFi and perhaps other unlicensed services, which prompted the addition to the standard even though that change in policy is only just going through in the US, with many other countries promising to follow suit during 2020. The WiFi Alliance trumpeted this development just before CES 2020, the last major industry show to go ahead before the Covid-19 shutdown, calling 6 GHz “an important portion of unlicensed spectrum that may soon be made available by regulators around the world.”

Many countries have now announced intentions to liberate some of this spectrum for unlicensed use, with the UK’s Ofcom for example in January 2020 announcing its proposal to make 500MHz of that spectrum in the 6GHz frequency band available for WiFi 6/6E. Then Europe’s Electronic Communications Committee (ECC), which is separate from the EU and represents instead the larger number of 48 European countries belonging to CEPT (European Conference of Postal and Telecommunications Administrations), is working to harmonize spectrum in this band with the aim of enabling first deployments before the end of 2020.

It is worth remembering that the old distinctions in emphasis between cellular and WiFi will still exist in the new era, even if there is more convergence between them, which we will come back to shortly. In fact, while 5G still offers large area coverage and high speed mobility within the whole supported area, WiFi 6/6E still does neither of those and is not intended to do so. Conversely, WiFi will still work out cheaper for devices such as tablets and smart TVs that are either fixed, or if roaming generally only access streaming video over WiFi. The downside of 5G for them is that it not only adds the cost of an additional radio but more significantly requires management of SIM cards and mobile subscriptions. That is why cellular enabled tablets never took off and we do not anticipate 5G changing that.

Partly for that reason we do anticipate some convergence between cellular and WiFi in the 5G era. This is why the WBA came together with the Next Generation Mobile Networks Alliance (NGNA), an association of mobile operators, vendors, manufacturers and research institutes focused on mobile broadband, late in 2019 to produce a new whitepaper outlining opportunities afforded by converged networks, arguing that WiFi 6E would play a critical role in the success of 5G.

The main point of the paper was to drive a renewed push towards heterogenous networks combining cellular and WiFi with automatic and transparent handover between the two. Although 5G now makes provision for indoor coverage with its inclusion of lower frequency spectrum below 1GHz, it is acknowledged that WiFi is better optimized for high capacity there within its bands.

In fact, 5G operates in two distinct frequency ranges, the first being Frequency range 1 (FR1) from 450MHz to 6GHz, which includes the LTE frequency range. Then frequency range 2 (FR2) is from 24.25GHz to 52.6GHz in what is called somewhat misleadingly the millimeter wave (mmWave) band. In fact, a mmWave travels at a frequency of about 300GHz, so even a 60 GHz band equates to a 5mm wavelength. But crucially 5G will not operate in those bands reserved for WiFi, that is around 2.4GHz, 5GHz and now 6 GHz, which underpin its high capacity.

A crucial addition to WiFi 6 was support for OFDMA (Orthogonal Frequency Division Multiple Access). This is an extension of the OFDM (Orthogonal Frequency Division Multiplexing) already used in earlier versions of WiFi. Here, OFDM splits the 20 MHz channel used typically in WiFi into multiple sub-carriers each of narrower band, often 64, each carrying less data but aggregating to more than a single whole channel.

This is because the sub-carriers are arranged such that interference between neighboring ones cancels out, avoiding the need for spacing via so-called guard intervals. This is effectively a form of parallel transmission where a data stream is split into multiple components, each carried separately but then reassembled at the destination. OFDM at least doubles capacity over wideband modulation across the whole channel.

OFDMA is then an extension of OFDM, allowing multiple clients with varying bandwidth requirements to be connected to a single WiFi AP (Access Point) simultaneously. This is done by allowing those narrowband sub-carriers to be allocated independently to multiple data streams, in effect partitioning a single 20 MHz channel into different sub-channels, known as Resource Units (RUs). Each AP can then communicate with multiple clients by assigning them to specific RUs. Furthermore, by subdividing the channel, applications that use small frames, as in video or audio streaming, can be transmitted to multiple endpoints simultaneously within a whole channel. This improves performance, reducing overhead and congestion within the WiFi domain.

OFDMA brought some challenges, above all that the benefits of OFDM in providing greater resistance against fading of the signal that can occur at some frequencies but not others can be lost if a given user only has a few subcarriers assigned. This would apply even more if the same sub-carrier was used to carry every single data symbol, in this case a pulse encoding a given number of bits. This has been solved through several measures, one being adaptive sub-carrier assignment where fast feedback about the channel is used to identify where fading is occurring and switching sub-carriers accordingly. Another is application of sub-carrier frequency hopping without such feedback to provide a statistical defense against signal fading.

The upshot is that with the benefits of OFDMA, that joint WBA/NGNA paper now argues convergence between 5G and WiFi 6 in the RAN (Radio Access Network) is essential for ensuring ubiquitous wireless coverage in dense locations and indoor deployments. The conclusion is that services to enterprises as well as consumers will be transformed by a combination of WiFi APs, 5G and also femtocells to enable ubiquitous connectivity.