# Enhancing 802.11ac MU-MIMO Efficiency: Simultaneous STA-AP and AP-STA Transmission Impact on Goodput Prediction

In 802.11ac' MU-MIMO two things (it's a XOR) can happen since wireless is half duplex: 1)at the same time more STAs transmit to AP 2)at the same time AP transmits to more STAs. Is this true? If so, if I want to predict max goodput (TMT).

If I have a 4x4 802.11ac AP so which supports MU-MIMO, I realize that if I connect two STAs (A and B) to WiFi they have a capacity link of 867Mbps both (I see this in the admin router pane. and I also get it from iw dev in the voice tx maximum bitrate).

If A wants to transmit B at full speed, how to predict/calculate the maximum theoretical througput (TMT)?

I thought about computing TMT=(1/2)*50%*C where

• 1/2 is for the nature of wifi, in which for A be able to talk to B, A needs to transmit to AP which (when A is not talking) transmits to B, so it's a cut of the speed because speed needs to be shared among A and AP. So I think I should put this.
• 50% is the 802.11 efficiency which can be approximately seen as this value, since packets like ACK,RTS,CTS are transmitted at very low speed so it's a cut of throughput
• C, the channel capacity. I think it should be 867Mbps. Which is the "dedicated" STA capacity that I see in the admin router panel and I also get it from iw dev in the voice tx maximum bitrate.

But the prediction, after having performed calculation, appears to be more "pessimistic" than the reality, after have performed an iperf3 test.

So, what am I doing wrong?

Can it be that 867Mbps is not the channel capacity but it's more like the "already computed" TMT? Since it's "dedicated" to STAs? Maybe the channel capacity is the aggregate sum of all the speed of each STAs can talk in every second to the AP?

• What is the point of your exercise? Usually, the theoretical, best-case throughput isn't achieved in practice and it's much more informative if you run a benchmarking tool in your exact environment.
– Zac67
Commented May 9 at 17:59