Transmission power and data rates

Hello guys, today I have an experience that made me think about a common misconception we as a wireless community have when adjusting our RF for green and brownfields, this post will be regarding how the data rate change really affects our RF designs.

To start on this let me state that I thought this was correct for a while and I think were the myth started, with this kind of images (credits to the owner):

Credits of the image to whom it might belong to. I have seen it several sites

The image, as you might already know, represents how the data rates get demodulated according to the physical distance, it exemplifies how the closer it is to the transmitter the easier it might demodulate a higher data rate and to be honest it makes perfect sense, it basically shows the physical range according to were the that data rate can work, right?

So, where is the misconception? I guess as me, you might have noticed forums, recommendations, and other places where it’s casually mention that increasing the minimum data rate to let’s say 24Mbps or 36Mbps will decrease the cell size and stop the interference; well that’s what I want to talk about.

 Transmit power

First, let’s do a refresh on the Co-channel interference concept. The interference occurs when there are devices close enough to “hear” each other and operate on the same channel, when this happens is not uncommon to see the to see the two devices contend for the medium access, this can also be called Co-channel contention (CCC).  

To mitigate or even avoid this issue there are two solutions, the obvious but difficult one: move to another channel; difficult because it will be so hard in today world to find a Wi-Fi channel that does not have another devide using it, maybe 6GHz?

The second solution is to plan and reduce your cell size; now just for you to be aware, there are two types of “boundaries” on a Wi-Fi cell, the association boundary and the CCI boundary.

Let’s start with the CCI, this boundary exists as the area where the listening stations (STA) that are or not part of the cell must be silent, this because they detected a signal over the minimum threshold; this behavior comes from the CSMA/CA that we have mentioned before, therefore this boundary can be won't necesarily be affected by low data rate signals, even when the lower data rates are disabled.

Important note: Remember that just because a signal cannot be decodified or processed by the STAs does not mean that a signal cannot be detectedor hear if you willl, this initiates the backoff timer to transmit, we have talked about this too before.

Give a check to my QoS out of the box post!

 

Now, the association or data rate boundary, this one work depending on the data rates enabled on the cell; basically, if you disable slower rates as 6Mbps or even 12Mbps, the physical range where the signal can be decoded decreases an it gives you the feel that your coverage is smaller; is it true that it is smaller? Well yeah! You won’t be able to "work" on that cell if you walk further away from where the signal can be decoded, this can be easily checked walking to the edge of your deployment, verify your coverage, then disable lower data rates and try again.  But, here is the misconception, as stated before, even while you can’t connect on that cell that does not mean that the signal is "heard" and triggers the backoff timers.

How does it look, here is an image from CWDP-303 book to make it easier to understand:

Ok so I gave you something to think about but not a solution, right? That doesn’t sound like me at all… I do know a way to make the CCI boundary smaller; the old fashion way is to make your cell smaller limiting the TxPower on your infrastructure devices, that will determine how far away you will be able to see the signal over your threshold signal.

Other ways to mitigate it is to increase the threshold but for this there is a small inconvenience; you can’t control the threshold of all the client devices, nevertheless I want to talk about it because from this idea there is one of the enhancements of Wi-Fi6, BSS coloring; while it will not be discussed in depth on this post I can tell you that basically with coloring you will be able to increase that threshold and ignore other signals as if they were not part of your deployment due to it’s “color”, of course to make it work properly a full Wi-Fi deployment it’s recommended and that can be expensive.

Note: For the CCI boundary we can’t overlook that the client has its own boundary and needs to be taken into account when doing the channel planning design.

 Data rates

As a small P.S. I wanted to talk about data rates, in this post so far, we have talked about how disabling them won’t decrease the CCI boundary but the association boundaries, does that mean it doesn't matter and we should enable all the lower rates? Not really.

There are many things that depend on the data rates and the cell lower-highest mandatory data rate, as beacons, multicast, sometimes I have seen roaming been affected by it. So, while it can’t fix the co-channel interference there must be a balance between the transmit power and how you don’t want that the beacons hog the medium with a slower data rate since it will take longer to send management frames.

My recommendation is before changing them always double check what are you designing for, voice, video, data. Keep in mind that the end application and the end devices will help your design and the more you keep those in mind the more accurate it will be, check the data sheet of your older devices, or for your special IOT devices, that’s a good guide.

I usually leave the lowest rate enable to be 12 or 24 on most deployments but I have used 36 too as the lowest mandatory rate on high density deployments with directional antennas, again always think on your use case.

Now a fun fact that I have seen it been overlooked, take a careful look at the lowest rates on the following image:

Credits of the image to whom it might belong to. I have seen it several sites

If you don’t know the index about acronyms on the top of the table, I will leave it at the end of the post, right now focus on the data rates; can you see how MCS 0 can be as low as 6.5 or 7.2 Mbps? Newer protocols use MCS data rates and those need to be enabled or disabled too for some use cases so always keep in mind not to only check the legacy rates but the newer too.

Am I saying you should always disable MCS 0? No, I’ll be honest, historically I have seen a lot of cases where a bad client-side driver can mess up your communication, that was sometime ago and should be better now, the point here is try to keep these new 802.11 revision in account too. 

Thanks for reading.

Dan Lopez

 

And as promise here is a quick reference for the data rate table I loved how simple this page explain it so take it with you if you want to learn more 

Reference articule: Panagiotis Vouzis, "What is the MCS Index?", from netbeez.net,  September 16, 2020

Reference book: CWDP-303 Certified Wireless Design Professional: Official Study Guide by Tom Carpenter. Certitrek publisher.