A-MPDU aggregation for 802.11ac

[ANDRÉ JÖNSSON]

Hello everyone! 

I'am trying to test the A-MPDU aggregation for 802.11ac. And as I understand it i can just change the default value MaxAmpduSize in AmpuStandardAggregator from 65535 to approximately 4MB?

When i run the simulation with a data rate of 150Mbps with 40MHz i get a throughput of 134 Mbit/s. I think this is a bit to high? Does anyone know what iam doing wrong. I attached the script that i am running.  

Best regards André

[Sebastien Deronne]

802.11ac? Are you serious or not? You just wrote it in the script: wifi.SetStandard (WIFI_PHY_STANDARD_80211n_5GHZ); => it's 802.11n!!!
802.11ac is only under development so far.

And yes, with a very large aggregation size you may get really close to the PHY bitrate in the best conditions.

 uint32_t nMpdus = 2900 ==> it is quite useless to use such HUGE aggregation size, since the standard anyway limits the maximum frame duration!

[ANDRÉ JÖNSSON]

Hello Sebastien! 

Sorry, maybe i was not clear enough. I know that i am using 802.11n and that was what i tested. And my real question was if if the value was correct. Never mind about 802.11ac, it was me that was unclear. But we are working on adding features for 802.11ac so thats why i mixed it up :). I also have one question about the ht-wifi-network. There we think that the values are pretty low. we got the following rates: 

DataRate Throughput

6.5 5.53472

13 10.0072

19.5 13.6825

26 16.8126

39 21.6466

52 25.4197

58.5 27.0718

65 28.3472

7.2 6.07642

14.4 10.9116

21.7 14.8908

28.9 18.1138

43.3 23.1128

57.8 26.967

65 28.5827

72.2 29.884

13.5 10.4253

27 17.5133

40.5 22.6818

54 26.5407

81 32.0131

108 35.9404

121.5 37.3982

135 38.8361

15 11.3627

30 18.8569

45 24.1443

60 28.0952

90 33.584

120 37.4559

135 38.849

150 40.2621

Sorry about the mix-up, regards André

[Sebastien Deronne]

Those values seems correct when aggregation is not used.

Note that 802.11ac is under development by Ghada Badawy and myself.
This does not mean you cannot work with us, everyone is welcome.

[ANDRÉ JÖNSSON]

Thanks a lot Sebastien, I really appreciate it!

Regards André 

[zakk...@gmail.com]

Hi Sebastian,

When is 802.11ac is expected to be available? Would it support MIMO and the rate adaptation? 

802.11n do not support multiple antennas and the rate adaptation (Minstrel works only up to 24 Mbps).  Showing results with 802.11n without MIMO is quite limited although all the main MAC features are implemented. 802.11a is more stable but it is  kind of obsolete.

Are there any plans to include MIMO and fix the rate adaptation 802.11n model as well?  If they are already implemented could we access the files before official release?

The other problem with the WiFi model is fading. I would like to use my own traces or an 802.11n or 802.11ac channel models for specific use cases.Are there any plans to enable that? Could you give me any guidelines how to insert fading traces to WIFI model?

Thanks for all of your efforts!

Aly

[Sebastien Deronne]

Roughly, I expect to deliver it for ns-3.24 (around September), but it may be postponed if it's not finished by that time.

No, MIMO is an other feature to be implemented for both 802.11n and 802.11ac.

Same comment for rate adaptation.

What do you mean by fading traces?

[zakk...@gmail.com]

Hi Sebastian,

I want to be able to model the small scale fading beyond just using the Rayleigh by tuning the Nakagami. LTE model allows to insert channel files which are generated with MATLAB.

https://www.nsnam.org/doxygen/trace-fading-loss-model_8cc_source.html

Could something like this be implemented for the WiFi model as well?

MIMO is of the  main improvement for 802.11n and 802.11ac.  I wished NS3 would have that capability in the near future too.

Thanks,

Aly

[Sebastien Deronne]

This does not yet exist for wifi channels, but you may have a try to implement it by yourself.

Contributors are always more than welcome :-)