[Bloat] ARAMIS: impressive wifi rate and channel width control algorithm for MIMO networks

[Bloat] ARAMIS: impressive wifi rate and channel width control algorithm for MIMO networks

[Dave Taht]

http://cs.ucsb.edu/~laradeek/Secon13.pdf

I just met Lara. I am impressed.

-- 
Dave Täht

Re: [Bloat] ARAMIS: impressive wifi rate and channel width control algorithm for MIMO networks

[Juliusz Chroboczek]

> http://cs.ucsb.edu/~laradeek/Secon13.pdf

Do you understand why ARAMIS with training performs so much better than 
Minstrel in the
20 MHz interferer case?

--jch

Re: [Bloat] ARAMIS: impressive wifi rate and channel width control algorithm for MIMO networks

[Felix Fietkau]

On 2013-12-10 09:33, Juliusz Chroboczek wrote:
>> http://cs.ucsb.edu/~laradeek/Secon13.pdf
> 
> Do you understand why ARAMIS with training performs so much better than 
> Minstrel in the
> 20 MHz interferer case?
One reason could be that the tests were made using the ancient 2.6.32
kernel. I've added many improvements and fixes to minstrel_ht since that
release.

- Felix

Re: [Bloat] ARAMIS: impressive wifi rate and channel width control algorithm for MIMO networks

[Thomas Hühn]

[-- Attachment #1: Type: text/plain, Size: 1219 bytes --]


Am 11.12.2013 um 14:40 schrieb Felix Fietkau <nbd@openwrt.org>:

> On 2013-12-10 09:33, Juliusz Chroboczek wrote:
>>> http://cs.ucsb.edu/~laradeek/Secon13.pdf
>> 
>> Do you understand why ARAMIS with training performs so much better than 
>> Minstrel in the
>> 20 MHz interferer case?
> One reason could be that the tests were made using the ancient 2.6.32
> kernel. I've added many improvements and fixes to minstrel_ht since that
> release.
> 

Iw as quite surprised why certain system research publications do still not provide sufficient information that enables someone to repeat their experiments. What version of the mac80211 subsystem did they used ? The reference just states: [6] Minstrel HT Linux Wireless, ”http://linuxwireless.org/en/developers”
Unfortunately there are many system relevant papers within networking that do not provide sufficient citations and experiment descriptions about software version (e.g. GIT hash), date, ect.
It appears to me that also reviewers do not put enough value to that. 

Thomas



> - Felix
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat


[-- Attachment #2: Type: text/html, Size: 2041 bytes --]