This topic is to discuss the following lesson:
Hello Rene,
a question regarding the channel re-use thing: Is WiFi 6 also considered a valid solution for high density WiFi? I at least think so, since WiFi 6 should have more non-overlapping channels, right?
Thanks in advance!
Kind regards,
Mirko
Hello Mirko
WiFi 6 is also considered a valid solution for high density WiFi. WiFi 6 has two specific features that aid in increasing capacity: Spectral efficiency, as well as an extension of the available frequencies and channels.
Spectral efficiency refes to the methoology by which multiple clients are managed and how efficiently the available spectrum is leveraged. WiFi6 uses what is known as orthgonal frequency division multiple access (OFDMA) which brings better spectrum use, improved power control to avoid interference, and enhancements like 1024‑QAM, MIMO and MU-MIMO for faster speeds.
WiFi6E, where “E” stands for “Extended”, is the version of WiFi6 that includes the extension of the WiFi 6 standard into the 6 GHz band. This new band opens up more than 1,200 MHz of spectrum for WiFi which means we have more channels available to us than those provided by the 2.4 and 5GHz bands alone. This does indeed provide us with an additional set of non overlapping channels, but only for clients that support this extended frequency range.
I hope this has been helpful!
Laz
Hello.
I don’t think I quite understand what bandwidth means in the context of wireless networks. Like the air has a certain amount of data that it can transmit per second?
With channel bonding, if you combine two 20 MHz channels into one 40 MHz one, does the bandwidth simply increase because the device has more frequences that it can use to transmit that data without any interference?
What if I have just one BSS? Couldn’t I combine every single 5 GHz channel into one? 
Thank you.
David
Hello David
With a wired connection, the bandwidth stated is quite straightforward. You will typically be able to reach very close to the stated bandwidth (i.e. 100 Mbps, 1000 Mbps, 10 Gbps etc) on such a link if there’s no congestion or if the resources of the router or switch are not reaching their limits. However, for a wireless link, there are many more factors involved.
Strictly speaking, if you have only one transmitter and one receiver, and you have no other wireless interference in the area, the bandwidth available between the sender and receiver is dependent only on the standard being used. For example, IEEE 802.11g gives you up to 54 Mbps, 802.11ac gives you up to 6.933 Gbps and so on. Now these as you know are theoretical, but they’re based on the specific parameters of the standard being used (encoding, frequencies, channel widths, modulation and others). So it’s not a matter of the “air” but of the transmitter’s and receiver’s capabilities.
Other factors that will affect the real achievable bandwidth include interference, obstacles, client density, AP capabilities and many more, as you have already seen from this and other lessons.
In wireless networks like Wi-Fi, various modulation mechanisms are used to encode data onto wireless waveforms. One that is used in many Wi-Fi standards is Orthogonal Frequency Division Multiplexing or OFDM. On a wireless channel with a channel width of 20 MHz OFDM creates what are known as subcarriers. A subcarrier in OFDM is a smaller frequency slice within the main channel bandwidth, individually modulated to carry part of the overall data, and precisely spaced to avoid interference with neighboring subcarriers.
A 20 MHz channel width using OFDM has a total of 64 subcarriers, 52 of which are used to carry data. The other 12 are used for control and as “guards” on either end of the channel to avoid interference with neighboring frequencies. Theoretically, by doubling the channel width, you can double the number of subcarriers. Actually, you get a larger number of data-carrying subcarriers because for a 40MHz channel width, you still only need 12 subcarriers as non-data-carrying subchannels, so you get more than double the number of data carrying subchannels, i.e. 114.
BUT, there is a tradeoff that doesn’t allow you to get twice as much throughput. Wider channels mean:
- they are prone to more interference noise and congestion, meaning in practice, they may actually be slower in a “crowded” wireless environment
- fewer non overlapping channels reduces the number of non interfering communications
- in a particularly crowded environment, you may find that using smaller channel widths actually allows a deployment to perform better than having channels bonded.
You could, if you have only a single client and a single AP. But this is rarely the case. Even if it is, you will definitely have interference from nearby sources. And you also must ensure that your client and AP both support the level of channel bonding you want to achieve.
So there is a tradeoff, and you must examine each case independently.
I hope this has been helpful!
Laz
Hello, everyone.
Regarding High Client density, is using a multi-cell design (one AP broadcasts multiple cells, say one for 2.4, one for 5 and one for 6 GHz) also a good solution to handle a lot of clients? Or broadcast the same SSID but on different channels.
Also, channel bonding isn’t really useful in high density scenarios, or is it? Sure, you can transmit on more frequencies and have a higher throughput but when we’re talking about a high density deployment, we aren’t only concerned about the throughput but also about the reliability or the general air and channel utilization.
If you have a few devices, then sure, channel bonding can help. But if you have 200 of them, they will all crowd the space and compete for airtime. If a device does get its turn then sure, it can transmit a massive amount of data but only for whatever its time duration is during its turn. Not to mention that not everyone has to transmit a lot of data.
The point that I thought about above also reduces the amount of channels that you can use which could be useful in a high density deployment.
It feels like it all boils down to many different things. We need coverage, we need a good speed/throughput, we need the density to be distributed so all clients have equivalent access to the channel, it gets pretty complex.
//Edit: Without getting too deep into this, is OFDM the opposite of channel bonding in a way? Channel bonding combines channels to achieve a higher frequency range therefore there is more throughput. OFDM takes those 20 MHz and divides them into subcarriers. Then, for example, your wireless devices could get a slice of the channel without interfering with the rest. So you divide a channel, get less throughput, but support more stations at the same time.
Or is this OFDMA?
Thank you.
David
Hello David
You’ve discussed some fascinating aspects of Wi-Fi design for high-density environments. The bottom line is that each technique that you can use to increase capacity has a trade-off. You improve in one area, but you need to deal with another…
For example, concerning your suggestion here:
This is definitely a good approach, since you are dividing the clients among three completely isolated frequency ranges. This however results in very different capabilities given to each of these groups of clients. The 2.4GHz range is quite small (3 overlapping channels) and is prone to more interference than the other ranges, so you’re giving different capabilities to different groups. It’s not necessarily bad, it’s just something to keep in mind.
No it isn’t. One fundamental technique for high-density environments is channel reuse, and if you use channel bonding, you are reducing the number of available nonoverlapping channels, so there’s a direct tradeoff.
That’s exactly correct. But knowing what factors affect the network in what way is an important part of the design process.
Hmm, it may seem that way but not quite. OFDM does divide a single channel (e.g., 20 MHz) into subcarriers to transmit data in parallel. While efficient, it doesn’t inherently support multi-user communication. The same user uses those subcarriers.
However, OFDMA used in Wi-Fi 6 extends OFDM by assigning subsets of subcarriers to different users simultaneously. This reduces contention and improves efficiency in high-density scenarios as it more granularly distributes portions of the spectrum to different users. However, this isn’t the opposite of channel bonding, it actually can work with channel bonding.
Keep in mind that channel bonding increases bandwidth per client, while OFDMA improves multi-user efficiency. They can coexist but require careful planning.
I hope this has been helpful!
Laz