Network Field Notes
 Academy
AcademySep 14, 2026 · 8 min

Bands, Channels and Widths

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 WIRELESS ACADEMY, LESSON 2 Every doubling of width halves the number of cells you can have. 20 MHz 25 channels 40 MHz 12 channels 80 MHz 6 channels 160 MHz 2 channels, both DFS no DFS DFS, shared with radar US channels in 5 GHz. Counting them is the design decision; the width is just how you spend them.
Level 1

Fit the Band

0 points

placed, clear overlapping another AP DFS, radar can take it
Where this came fromWireless Academy, lesson 2. The trade nobody feels until they run out of channels

Last week was one radio and one number. This week is the room that radio has to live in, and the first real trade in wireless design.

Three words, in plain language. A band is a stretch of spectrum a regulator has set aside. A channel is a slice of that band your radio actually sits on. Width is how big a slice you took. That is the whole vocabulary, and the entire lesson is what happens when you take a bigger slice.

What you actually have

In the US there are three bands worth your time, and they are wildly different sizes.

2.4 GHz gives you eleven channels, and only three of them fit side by side without overlapping: 1, 6 and 11. The rest overlap their neighbours, which is why channel 3 is not a clever compromise, it is two problems at once. Everything else lives here too, so the band is loud before you arrive.

5 GHz is the working band, and it comes in four blocks you should know by name. U-NII-1 is channels 36 to 48. U-NII-2A is 52 to 64. U-NII-2C is 100 to 144, the big one. U-NII-3 is 149 to 165. Twenty five channels of 20 MHz, which sounds like plenty until you start widening them.

The catch is DFS. Everything from 52 to 144 is shared with radar, mostly weather and military, and the rules are not negotiable: your AP has to listen before it transmits there, and if it hears a radar pattern it must leave that channel and stay off it for thirty minutes. That is real capacity, and you should use it, but it is borrowed rather than owned.

6 GHz is the new room, and it is enormous: 5.925 to 7.125 GHz, fifty nine 20 MHz channels, and no DFS anywhere in it. Seven of those channels are 160 MHz wide, which is six more than 5 GHz can offer. Outdoors at standard power you need AFC to tell you what you may use, and only Wi-Fi 6E and Wi-Fi 7 clients can see any of it at all.

The number that matters

Doubling the width doubles the number of subcarriers, so it roughly doubles the data rate. Nobody argues with that half. Here is the other half.

Doubling the width raises the noise floor by 3 dB. A wider channel is a bigger bucket, and it collects proportionally more noise. Your radio needs 3 dB more signal to hold the same modulation, so the cell gets smaller. Go from 20 to 80 MHz, four times the width, and you have given away 6 dB of link budget. That is the same 6 dB that costs you half your distance, from lesson one.

And the part people feel last: every doubling of width halves how many non-overlapping channels you have. In 5 GHz that takes twenty five channels down to six at 80 MHz, and down to two at 160 MHz, both of which are DFS. A design is not a single cell, it is cells next to each other, and cells next to each other need different channels.

So the trade is not "wider is faster." The trade is wider is faster per cell, smaller per cell, and fewer cells. In a warehouse with four APs, fine. In a lecture hall with thirty, it is a way to build one slow network out of a lot of expensive hardware.

Play it, it is faster than reading it

Drop the access points onto legal channels. The rules are the real ones: 40 MHz has to sit on a proper pair, 160 has to sit on a proper eight, DFS is shaded, and anything that overlaps another AP turns orange. Level two is the lesson.

What the gear shows you

Width is a radio setting, not an SSID setting, and that trips people up on both platforms.

In Aruba Central, width lives in the RF profile alongside the channel list and transmit power, and nowhere in the WLAN wizard. The important part is that the profile is an object you assign rather than a switch you flip: you choose which APs get which profile, so the group is a convenient default and not the only unit, and a single AP can be handed its own. An override on one AP beats what the group says. That is how you fix one bad room, and it is also how somebody's fix from eight months ago is still quietly in force on an AP nobody has looked at since.

To see what an AP actually ended up on, rather than what you asked for, show ap bss-table gives you the BSSID, the channel and the width in use. That column is where you find out your 80 MHz plan quietly became 40 because a neighbour was already sitting there.

In Mist, width is in the RF template at org level, and you can override it on a single AP from that AP's page when one room genuinely needs something different. Client insights then shows you the PHY rate that resulted. Make the decision once in the template, and treat each per AP override as something that owes you an explanation later.

On the Sidekick, the channel view shows you the width as occupied spectrum rather than as a setting, which is the only view that tells you what your neighbours chose.

The lab

One AP, one client, about forty minutes.

  1. Put the AP on a 20 MHz channel in U-NII-1, somewhere clear, and connect one client a couple of metres away.
  2. Note three things: the client's PHY rate, its RSSI, and the AP's airtime utilisation.
  3. Run a file transfer big enough to last a minute and record the throughput.
  4. Change the radio to 40 MHz, let the client reconnect, and repeat every measurement.
  5. Do it again at 80 MHz.
  6. Now walk to the edge of coverage, where RSSI is around -70 dBm, and repeat at 20 and at 80.

What you should see. Here is what the model behind the simulator on this site says, for a Wi-Fi 6 client with two spatial streams, a 0.8 microsecond guard interval, a 7 dB noise figure, and 1500 byte frames aggregated 64 deep with the medium to itself.

Width Close in, RSSI -55 dBm At the edge, RSSI -72 dBm
20 MHz MCS 11, 287 Mb/s PHY, about 235 Mb/s of throughput MCS 5, 138 Mb/s PHY, about 117 Mb/s
40 MHz MCS 10, 516 Mb/s PHY, about 399 Mb/s MCS 4, 206 Mb/s PHY, about 172 Mb/s
80 MHz MCS 9, 961 Mb/s PHY, about 673 Mb/s MCS 3, 288 Mb/s PHY, about 236 Mb/s

Two things in that table are worth more than the numbers themselves.

The first is that throughput never keeps up with the PHY rate, and the gap widens as the rate climbs: 82 percent of the PHY rate at 20 MHz close in, 70 percent at 80 MHz. Preamble, interframe spaces, backoff and the acknowledgement do not get smaller when the data gets faster, so the faster you go, the larger a share of your airtime is spent not carrying data. That is lesson 5 arriving early.

The second is the one I had wrong before I ran the numbers. Wider is still faster at the edge. At -72 dBm the 80 MHz link sits three MCS steps lower than the 20 MHz link and still moves twice the data, because four times the subcarriers beats a few steps of modulation. The cost of width is not that it is slower out there. The cost is where the link gives up entirely: in this model a 20 MHz link holds on to about -91 dBm and an 80 MHz link quits around -85, six decibels earlier, which by the rule from lesson one is half the distance. Width does not make the far edge slow, it moves the far edge closer.

And none of that counts the cost the game shows you, which is that you have fewer cells to place. In a real room the noise floor is not thermal either, it is other people, and a wider channel scoops up more of them.

A word on where these numbers come from. They are computed, not measured: the same MCS tables, airtime and noise floor model that drives the simulator, run for this lesson. They are here to make the shape of the trade concrete rather than to tell you what your gear will print. When I have run this lab on my own AP I will replace them with measured figures and say where the model was wrong, because it will be wrong somewhere, and the gap between the two is usually the most interesting part of the exercise.

What means it is broken. If the PHY rate does not move at all when you change width, the client never reconnected, or it is a 20 MHz only device, and plenty of scanners and handhelds are. If throughput goes up but airtime goes up faster, you are watching retries, not capacity.

Three questions

  1. Your site has fourteen APs in one open floor and the customer wants 80 MHz everywhere in 5 GHz. How many non-overlapping channels do you have without DFS, and what is the actual reuse distance you are asking for?
  2. An AP on channel 100 stops serving clients for half an hour and then comes back on its own. What happened, and which band would have avoided it?
  3. Two APs are on 40 MHz channels that share one 20 MHz half. Is that better or worse than putting them both on the same 40 MHz channel, and why?

Next week

Lesson 3 is the link budget: EIRP, antenna gain, receive sensitivity and free space path loss, and predicting the RSSI at 10 m before you go and measure it.

Reading path

Wireless Academy part 2 of 2

1What a Radio Actually Sends2Bands, Channels and WidthsYou are here

Comments

    Comments appear straight away. Your email, if you leave one, is stored so I can reply to you and is never shown on the page or passed to anyone. Anything abusive gets removed.

    More field notes

    OlderConverting a Mesh Over the Air Strands Every PointNewerI Held the Mesh Planner Against a Real Point and It Was Wrong About One Thing