I don't have any real answers for you. I'd look at the interframe spacing
and timing; see if it's due to big gaps in transmission/reception, see if
it's something silly like AMPDU density being way too big, or maybe just
lots of retries.

I'm sorry I can't be more helpful than this!



-adrian



On 11 October 2013 11:18, Ramanujan Seshadri <ksramanu...@gmail.com> wrote:

> It's in the NIC. But it seems that it sometimes aggregates more than the
> set BA win size (in the driver).  In fact i was wondering - if it can
> actually achieve up to 80 Mbps throughput with just 8 frames aggregated. In
> fact the aggregation size doesn't seem to matter much for the throughput
> because any aggregation from 5-8 seems to get the same throughput. So, i
> was wondering if this is something weird in NIC or can we do some thing to
> control it in the driver ?
>
> ram
>
>
> On Thu, Oct 10, 2013 at 4:41 PM, Adrian Chadd <adr...@freebsd.org> wrote:
>
>> I forget .. is the driver or the NIC itself doing aggregation?
>>
>>
>> -a
>>
>>
>> On 10 October 2013 10:33, Ramanujan Seshadri <ksramanu...@gmail.com>wrote:
>>
>>> Hello all,
>>>   i was testing some features of ralink rt2860 driver. i saw a bit weird
>>> results. below is the result of what i saw. The throughput achieved
>>> doesn't
>>> change much when the MPDU aggregation in an AMPDU goes from 3 to 8. And
>>> also, the driver doesn't seem to aggregate more than 8 MPDU's in an AMPDU
>>> even when the BA win size is set to max of 64. i was testing this by
>>> fixing
>>> the rate at MCS 15
>>>
>>>  Is there anything else i need to do to force the driver to aggregate
>>> more
>>> MPDU's in an AMPDU ?
>>>
>>>
>>>
>>> BA Win Size           No. of MPDU's            Throughput(Mbps)
>>>                                 aggregated
>>>
>>> 1. 64                             7-8                            85.7
>>>
>>> 2. 32                             6-7                            84.2
>>>
>>> 3. 16                             6-7                            81.7
>>>
>>> 4.  8                              6-7                            79.9
>>>
>>> 5.  4                               5                              84.7
>>>
>>> 6.  2                               3                              71.8
>>>
>>> 7.  1                               2                              73.5
>>>
>>>
>>> Thanks
>>> ram
>>> _______________________________________________
>>> freebsd-wireless@freebsd.org mailing list
>>> http://lists.freebsd.org/mailman/listinfo/freebsd-wireless
>>> To unsubscribe, send any mail to "
>>> freebsd-wireless-unsubscr...@freebsd.org"
>>>
>>
>>
>
_______________________________________________
freebsd-wireless@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-wireless
To unsubscribe, send any mail to "freebsd-wireless-unsubscr...@freebsd.org"

Reply via email to