Since the introduction of wireless to networking, the WLAN has been limited by one major factor: the radio spectrum. When Wi-Fi was first introduced, the spectrum was limited to less than 100Mhz in the 2.4GHz range. Throughout the various evolutions of wireless networking, two of the main goals have been to more efficiently use spectrum, and to add spectrum to overcome overlap and throughput constrictions. The new 1200MHz of spectrum provided by the 6GHz channels allows us to rethink how we design channel plans, plus leveraging larger channel widths – previously considered bad practice – allows us to overcome previous constraints.
The Overlap Problem
Wireless networks operate by using a spread spectrum concept; this means the energy from the transmission spans across a portion of the spectrum, in our case, either 22MHz or 20MHz wide. This is how we come up with the three non-overlapping channels in the 2.4GHz band. With 5GHz support in 802.11a/ac/ax, our usable spectrum increases 5x to 500 MHz, allowing for 25 non-overlapping 20MHz channels. Wi-Fi 6 introduces another frequency range for our use: 6GHz with up to 1200MHz of new spectrum (this varies by geography). If your region supports all 1200MHz of the new spectrum, 59 additional channels are available to client devices! As client devices that are capable of supporting this new spectrum come to market, at a minimum it is important to think about how our wireless designs will change in key high-density areas.
Capacity and Throughput
In 2021, the number of mobile devices almost reached 15 billion and is expected to soon reach 16 billion according to Statista research. With the increase in the number of devices needing to be connected, the amount of data we use during each session has exponentially grown. For example, using TikTok (a popular video sharing platform) for 5 minutes creates almost 100MB of data. Instagram, another popular video and photo-sharing platform, uses nearly 40MB of data in the same timeframe. On the extreme end, watching a 4K ultra high-definition video stream uses close to 6GB of data per hour, while browsing the web uses approximately 15MB per hour, according to our tests. We need to walk the fine line of capacity and throughput to provide the bandwidth for these data-hungry services.
With wired networking, if we needed additional capacity, one could easily add another switch, thus providing additional ports and a known throughput through the uplinks. With wireless networking on the other hand, it is not as easy. Adding an extra AP may not increase wireless capacity as much as you would think. Let’s look at it slightly differently. Imagine you oversee the designing and building of a complex system of roads to get people in and out of your town. The easiest way to accomplish this is to create a single-lane, bi-directional roadway. Traffic can easily flow in their respective lanes in and out of town. Now, fast-forward a few years, and you have a fantastic vibrant downtown center, thus increasing traffic on your single-lane road, which requires you to invest in a multi-lane roadway. Each of these lanes is equivalent to our channels in wireless. The more lanes we have, the more traffic we can support at any time. Once a lane is occupied, though, congestion can start to form. Taking this knowledge back to our original concept of adding APs, we see how if we add that AP to an already occupied channel, we don’t fully introduce additional capacity; it is a limited capacity.
Over the years, our data usage profiles have changed from email and Internet browsing to TikTok, Instagram, and Netflix. How we use these lanes has also evolved, growing from 3 to 12 to 25. But what happens when you run out of room to add lanes? 401 Highway in Ontario, Canada, is one of the busiest in the world. With over a dozen lanes, there can still be congestion, so what if we could fit more data (aka people) into a single vehicle and on a lane simultaneously?
“Adding an extra AP may not increase wireless capacity as much as you would think.“
These heavy data usage applications are oversized loads on a roadway. The average lane is around 3 – 4m in width, a known measurement, just like the bandwidth available in channel width, is a known theoretical maximum. Two lanes are required to clear a payload size of 6m in
width. We can accommodate that on the roads by allowing vehicles to straddle both lanes using pilot vehicles and traveling at non-congested times. With wireless, we adapt to these oversized loads by leveraging larger channel widths than 20MHz, such as 40MHz, 80MHz, and even 160MHz, which is done by combining multiple channels into a single channel. Sounds great. Well, there is a downside to this. Our 25 channels in 5GHz suddenly become 12, or even 2 at 160MHz! By combining the 25 channels in 5GHz, we effectively reduce our overall aggregate throughput and capacity. Because of this, more spectrum is needed for wireless devices and 6GHz solves that need.
6GHz to the Rescue! 6GHz represents the potential for 1200MHz worth of usable spectrum, depending on geographical region. We don’t necessarily need the 59 20Mhz channels or even the 29 40MHz channels. The benefit of 6GHz truly is the 14 80MHz or 7 160MHz channels available in addition to the 25 20MHz or 12 40MHz channels in the 5GHz band. Now we can easily direct high data usage devices, such as streaming and virtual reality devices, to associate solely to the 6GHz band, allowing 80MHz and 160MHz channel bandwidth, and letting them band roam to 5GHz when needed.
For the first time in a while, we can have a clean slate that will allow us to start with new design guidelines and no preconceived notions about the “recommended” way of doing things. Client devices are required to support WPA3 to operate in the 6GHz band. This requirement is a driving factor to deploy a new network name (SSID) that solely operates in the 6GHz band. There is no requirement for us to have to support older legacy devices!
To further assist with the allocation of the frequency, three new device classes have been created: Low Power Indoor (LPI) AP, Standard Power (SP) AP, and Very Low Power (VLP) AP.
LPI devices are fixed indoor-only APs and operate in such a way that they reduce their impact on incumbent services already operating within the 6GHz band. By limiting the EIRP (radiation power) to 30dBm at the AP and 24 dBm at the client, we will be able to utilize the 1200MHz spectrum at higher channel widths efficiently. LPI APs EIRP will be enforced by requiring permanently attached integrated antennas. This removes the ability to add a higher gain antenna and increases the EIRP above the maximum limits.
SP devices are primarily designed for indoor and outdoor usage but will operate in a subset of the 6GHz band, U-NII-5, and U-NII-7. Because these devices are authorized for outdoor use, the maximum EIRP is increased to 36 dBm. As this higher EIRP may interfere with existing incumbent users on the 6GHz band, here in the US, the FCC requires a spectrum management service to be used called Automatic Frequency Coordination (AFC). When an outdoor device comes online, it is necessary to communicate with a local AFC system to retrieve a list of allowed and prohibited frequencies using geolocation. While this is new to the Wi-Fi world, CBRS and other technologies have used this for some time already.
The final device class, VLP, is designed for transportation vehicles such as cars, trains, and others. The maximum EIRP for this class will be around 14 dBm. VLP could also be leveraged for high throughput personal area network devices such as VR headsets. Time will tell what happens with VLP, as the primary focus seems to be on LPI and SP devices.
Regardless of which class of device you deploy on 6GHz, one of the goals is to be better RF neighbors and coordinate usage while also allowing crosstalk amongst different device classes. Through these new device classes, we can finally see the real-world benefits of using 80Mhz and 160Mhz channel widths in the enterprise, not just at home!
The Need for Multi-Gigabit Ethernet
As we use these 80MHz and 160MHz channel widths, we cannot forget about our wired backhaul. The massive channels supporting these data-hungry applications are like the G4 Beijing-Hong Kong-Macau Expressway checkpoint that merges 50 lanes into 20!
As you evaluate new wireless equipment, ensure the Wi-Fi 6E APs you are looking at support either 802.3ad, Link Aggregation Control Protocol (LACP), or mGig, enabling support for either 2.5Gbps or 5Gbps. It is also worthwhile to pull two cables to every AP now, not just for the sake of redundancy, but to take advantage of LACP on mGig as well, enabling the possibility of dual 2.5Gbps or dual 5Gbps! Imagine pushing all that data traffic from your Wi-Fi 6 devices onto a single gigabit Ethernet connection. With Wi-Fi 6 it is possible to oversubscribe a 1Gbps Ethernet uplink.
We hope you enjoyed this look at Wi-Fi 6 channels and can appreciate that what we once considered the “norm” for channel design should be challenged with Wi-Fi 6. We have 1200MHz worth of new spectrum to set the tone for how we want to leverage it with our plans. Let’s push away from the past and embrace the bandwidth the 6GHz band brings to a new world of devices. And remember, as you design for these new challenges, the EtherScope nXG from NetAlly fully supports Wi-Fi 6 and the 6GHz band for analysis, and mGig Ethernet testing for end-to-end connectivity troubleshooting and validation. For wireless specialists, the AirCheck G3 Wireless Analyzer features all the power of the EtherScope nXG but without wired Ethernet testing (available Q4-2022).
Blake Krone is an independent Mobility Consultant and developer. His primary focus is providing solutions for the next generation of devices and business use cases for many Fortune 500 companies and startups. He has developed training materials and presentations through his experience deploying some of the largest single-site networks, sharing the knowledge and insights gained. When he isn’t designing and deploying networks, he builds data analysis tools and tests client devices and tools.