Dynamic Bandwidth in hw rate control

Dynamic Bandwidth in hw rate control

[Gaurang Ramesh Naik]

Hi,

I needed a small clarification related to dynamic bandwidth setup. The
documentation in wmi.h says " When enabled HW rate control tries
different bandwidths when retransmitting frames."

Does this mean that each packet is first always tried over VHT80 and
if that fails it goes to HT40 and then to HT20? I tried this with one
AP-Sta operating in VHT80 and one AP operating in its first secondary
channel. When I check the rx rate info using "iw dev wlanx station
dump", the bandwidth seems to sometimes alternate between 20 and 80,
but sometimes it stays fixed at 20. Hence, I was wondering what really
happens.

Thanks,
Gaurang.

_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

Re: Dynamic Bandwidth in hw rate control

[Ben Greear]

On 06/08/2016 08:10 PM, Gaurang Ramesh Naik wrote:
> Hi,
>
> I needed a small clarification related to dynamic bandwidth setup. The
> documentation in wmi.h says " When enabled HW rate control tries
> different bandwidths when retransmitting frames."
>
> Does this mean that each packet is first always tried over VHT80 and
> if that fails it goes to HT40 and then to HT20? I tried this with one
> AP-Sta operating in VHT80 and one AP operating in its first secondary
> channel. When I check the rx rate info using "iw dev wlanx station
> dump", the bandwidth seems to sometimes alternate between 20 and 80,
> but sometimes it stays fixed at 20. Hence, I was wondering what really
> happens.

For one thing, if the hardware detects interference on secondary channels,
then it may decide to transmit a 20Mhz encoding.

The rate-ctrl algs in the firmware differ from release to release, but
the general goal is to transmit with the fastest encoding rate.

Thanks,
Ben

>
> Thanks,
> Gaurang.
>
> _______________________________________________
> ath10k mailing list
> ath10k@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/ath10k
>

-- 
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc  http://www.candelatech.com

_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

Re: Dynamic Bandwidth in hw rate control

[Gaurang Ramesh Naik]

Hi Ben,

Thanks for your reply.

I wanted to know if the dynamic bandwidth option implemented in ath10k
is the same as that mentioned in the standard, or does it deviate? I
tried to look for any documentation on ath10k dynamic bandwidth, but
could not find any.

If I am not wrong, on the secondary channel, the sender senses the
channel for PIFS interval of time immediately preceding the start of
the TXOP (9.19.2.8 in the standard). In the static mode, if the
secondary channel is sensed busy, the sender waits until all channels
become idle; but in the dynamic mode, the sender sends on the primary
and (adjacent) secondary channel that are not busy.

However, the documentation for dynamic bandwidth in ath10k/mac.c seems
to suggest that the transmission retries are carried over narrower
bandwidth in dynamic mode. Does this mean the same as what is given in
the standard? Or is the dynamic mode implemented differently? Any help
is appreciated.

Thanks,
Gaurang.

On Thu, Jun 9, 2016 at 1:23 AM, Ben Greear <greearb@candelatech.com> wrote:
>
>
> On 06/08/2016 08:10 PM, Gaurang Ramesh Naik wrote:
>>
>> Hi,
>>
>> I needed a small clarification related to dynamic bandwidth setup. The
>> documentation in wmi.h says " When enabled HW rate control tries
>> different bandwidths when retransmitting frames."
>>
>> Does this mean that each packet is first always tried over VHT80 and
>> if that fails it goes to HT40 and then to HT20? I tried this with one
>> AP-Sta operating in VHT80 and one AP operating in its first secondary
>> channel. When I check the rx rate info using "iw dev wlanx station
>> dump", the bandwidth seems to sometimes alternate between 20 and 80,
>> but sometimes it stays fixed at 20. Hence, I was wondering what really
>> happens.
>
>
> For one thing, if the hardware detects interference on secondary channels,
> then it may decide to transmit a 20Mhz encoding.
>
> The rate-ctrl algs in the firmware differ from release to release, but
> the general goal is to transmit with the fastest encoding rate.
>
> Thanks,
> Ben
>
>>
>> Thanks,
>> Gaurang.
>>
>> _______________________________________________
>> ath10k mailing list
>> ath10k@lists.infradead.org
>> http://lists.infradead.org/mailman/listinfo/ath10k
>>
>
> --
> Ben Greear <greearb@candelatech.com>
> Candela Technologies Inc  http://www.candelatech.com

_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

Re: Dynamic Bandwidth in hw rate control

[Dave Taht]

On Thu, Jun 16, 2016 at 12:57 PM, Gaurang Ramesh Naik <gaurang@vt.edu> wrote:
> Hi Ben,
>
> Thanks for your reply.
>
> I wanted to know if the dynamic bandwidth option implemented in ath10k
> is the same as that mentioned in the standard, or does it deviate? I
> tried to look for any documentation on ath10k dynamic bandwidth, but
> could not find any.

So far as I know "rate control" is one of the major handwaves in the
standard and differs from vendor to vendor and 802.11 standard to
another.

The pieces of how 802.11ac should fall back to narrower channels,  I
think you've correctly identified below (as for width)...

but actually choosing a MCS rate, number of retries, formatting an
aggregate... and other factors, are... different.

I've tried to point at the first and best paper that shows how
minstrel works here:

http://blog.cerowrt.org/post/minstrel/

(and minstrel's known problems) but that's not what ath10k uses.

I would *welcome* full documentation (or code!) on how ath10k choses
its rates. Or any other wifi chipset's blob, frankly.

> If I am not wrong, on the secondary channel, the sender senses the
> channel for PIFS interval of time immediately preceding the start of
> the TXOP (9.19.2.8 in the standard). In the static mode, if the
> secondary channel is sensed busy, the sender waits until all channels
> become idle; but in the dynamic mode, the sender sends on the primary
> and (adjacent) secondary channel that are not busy.
>
> However, the documentation for dynamic bandwidth in ath10k/mac.c seems
> to suggest that the transmission retries are carried over narrower
> bandwidth in dynamic mode. Does this mean the same as what is given in
> the standard? Or is the dynamic mode implemented differently? Any help
> is appreciated.
>
> Thanks,
> Gaurang.
>
> On Thu, Jun 9, 2016 at 1:23 AM, Ben Greear <greearb@candelatech.com> wrote:
>>
>>
>> On 06/08/2016 08:10 PM, Gaurang Ramesh Naik wrote:
>>>
>>> Hi,
>>>
>>> I needed a small clarification related to dynamic bandwidth setup. The
>>> documentation in wmi.h says " When enabled HW rate control tries
>>> different bandwidths when retransmitting frames."
>>>
>>> Does this mean that each packet is first always tried over VHT80 and
>>> if that fails it goes to HT40 and then to HT20? I tried this with one
>>> AP-Sta operating in VHT80 and one AP operating in its first secondary
>>> channel. When I check the rx rate info using "iw dev wlanx station
>>> dump", the bandwidth seems to sometimes alternate between 20 and 80,
>>> but sometimes it stays fixed at 20. Hence, I was wondering what really
>>> happens.
>>
>>
>> For one thing, if the hardware detects interference on secondary channels,
>> then it may decide to transmit a 20Mhz encoding.
>>
>> The rate-ctrl algs in the firmware differ from release to release, but
>> the general goal is to transmit with the fastest encoding rate.
>>
>> Thanks,
>> Ben
>>
>>>
>>> Thanks,
>>> Gaurang.
>>>
>>> _______________________________________________
>>> ath10k mailing list
>>> ath10k@lists.infradead.org
>>> http://lists.infradead.org/mailman/listinfo/ath10k
>>>
>>
>> --
>> Ben Greear <greearb@candelatech.com>
>> Candela Technologies Inc  http://www.candelatech.com
>
> _______________________________________________
> ath10k mailing list
> ath10k@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/ath10k



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

Re: Dynamic Bandwidth in hw rate control

[Ben Greear]

On 06/16/2016 12:57 PM, Gaurang Ramesh Naik wrote:
> Hi Ben,
>
> Thanks for your reply.
>
> I wanted to know if the dynamic bandwidth option implemented in ath10k
> is the same as that mentioned in the standard, or does it deviate? I
> tried to look for any documentation on ath10k dynamic bandwidth, but
> could not find any.
>
> If I am not wrong, on the secondary channel, the sender senses the
> channel for PIFS interval of time immediately preceding the start of
> the TXOP (9.19.2.8 in the standard). In the static mode, if the
> secondary channel is sensed busy, the sender waits until all channels
> become idle; but in the dynamic mode, the sender sends on the primary
> and (adjacent) secondary channel that are not busy.
>
> However, the documentation for dynamic bandwidth in ath10k/mac.c seems
> to suggest that the transmission retries are carried over narrower
> bandwidth in dynamic mode. Does this mean the same as what is given in
> the standard? Or is the dynamic mode implemented differently? Any help
> is appreciated.

I don't know exactly how this works.  At least the sensing and stuff is probably
baked into the hardware.  What I notice is that the firmware passes a series
of rates to the hardware to use for each packet, and it includes different
bandwidth rates.

If you used my firmware and my kernel, you can see the (successful) transmit
rate for each frame.  Perhaps you could discover some info about how it works
in that manner if you could reliably add noise on an adjacent channel.

Upstream firmware can see tx rates using pktlog I think, but I have
never tried that.

Thanks,
Ben

>
> Thanks,
> Gaurang.
>
> On Thu, Jun 9, 2016 at 1:23 AM, Ben Greear <greearb@candelatech.com> wrote:
>>
>>
>> On 06/08/2016 08:10 PM, Gaurang Ramesh Naik wrote:
>>>
>>> Hi,
>>>
>>> I needed a small clarification related to dynamic bandwidth setup. The
>>> documentation in wmi.h says " When enabled HW rate control tries
>>> different bandwidths when retransmitting frames."
>>>
>>> Does this mean that each packet is first always tried over VHT80 and
>>> if that fails it goes to HT40 and then to HT20? I tried this with one
>>> AP-Sta operating in VHT80 and one AP operating in its first secondary
>>> channel. When I check the rx rate info using "iw dev wlanx station
>>> dump", the bandwidth seems to sometimes alternate between 20 and 80,
>>> but sometimes it stays fixed at 20. Hence, I was wondering what really
>>> happens.
>>
>>
>> For one thing, if the hardware detects interference on secondary channels,
>> then it may decide to transmit a 20Mhz encoding.
>>
>> The rate-ctrl algs in the firmware differ from release to release, but
>> the general goal is to transmit with the fastest encoding rate.
>>
>> Thanks,
>> Ben
>>
>>>
>>> Thanks,
>>> Gaurang.
>>>
>>> _______________________________________________
>>> ath10k mailing list
>>> ath10k@lists.infradead.org
>>> http://lists.infradead.org/mailman/listinfo/ath10k
>>>
>>
>> --
>> Ben Greear <greearb@candelatech.com>
>> Candela Technologies Inc  http://www.candelatech.com
>
> _______________________________________________
> ath10k mailing list
> ath10k@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/ath10k
>


-- 
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc  http://www.candelatech.com


_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

Re: Dynamic Bandwidth in hw rate control

[Gaurang Ramesh Naik]

Thanks Ben and Dave for your replies. So what I understand is that
dynamic bandwidth management is a part of the rate control algorithm
(which is implemented in firmware) and hence I have not much control
over how that works.

I am assuming your firmware is the one that can be downloaded from
http://www.candelatech.com/ath10k.php Is your kernel a modified one
too? When you say I can see the tx rate for each frame, is it through
some logs? I could try to get some info by using that. Thanks again.

Slightly unrelated, but is there a way I can play around with PIFS
value? I could not locate it in the ath10k source, so I am assuming it
is implemented in the firmware. I am trying to study some fairness
issues when legacy devices operate in the secondary channels.

On Thu, Jun 16, 2016 at 4:54 PM, Ben Greear <greearb@candelatech.com> wrote:
> On 06/16/2016 12:57 PM, Gaurang Ramesh Naik wrote:
>>
>> Hi Ben,
>>
>> Thanks for your reply.
>>
>> I wanted to know if the dynamic bandwidth option implemented in ath10k
>> is the same as that mentioned in the standard, or does it deviate? I
>> tried to look for any documentation on ath10k dynamic bandwidth, but
>> could not find any.
>>
>> If I am not wrong, on the secondary channel, the sender senses the
>> channel for PIFS interval of time immediately preceding the start of
>> the TXOP (9.19.2.8 in the standard). In the static mode, if the
>> secondary channel is sensed busy, the sender waits until all channels
>> become idle; but in the dynamic mode, the sender sends on the primary
>> and (adjacent) secondary channel that are not busy.
>>
>> However, the documentation for dynamic bandwidth in ath10k/mac.c seems
>> to suggest that the transmission retries are carried over narrower
>> bandwidth in dynamic mode. Does this mean the same as what is given in
>> the standard? Or is the dynamic mode implemented differently? Any help
>> is appreciated.
>
>
> I don't know exactly how this works.  At least the sensing and stuff is
> probably
> baked into the hardware.  What I notice is that the firmware passes a series
> of rates to the hardware to use for each packet, and it includes different
> bandwidth rates.
>
> If you used my firmware and my kernel, you can see the (successful) transmit
> rate for each frame.  Perhaps you could discover some info about how it
> works
> in that manner if you could reliably add noise on an adjacent channel.
>
> Upstream firmware can see tx rates using pktlog I think, but I have
> never tried that.
>
> Thanks,
> Ben
>
>
>>
>> Thanks,
>> Gaurang.
>>
>> On Thu, Jun 9, 2016 at 1:23 AM, Ben Greear <greearb@candelatech.com>
>> wrote:
>>>
>>>
>>>
>>> On 06/08/2016 08:10 PM, Gaurang Ramesh Naik wrote:
>>>>
>>>>
>>>> Hi,
>>>>
>>>> I needed a small clarification related to dynamic bandwidth setup. The
>>>> documentation in wmi.h says " When enabled HW rate control tries
>>>> different bandwidths when retransmitting frames."
>>>>
>>>> Does this mean that each packet is first always tried over VHT80 and
>>>> if that fails it goes to HT40 and then to HT20? I tried this with one
>>>> AP-Sta operating in VHT80 and one AP operating in its first secondary
>>>> channel. When I check the rx rate info using "iw dev wlanx station
>>>> dump", the bandwidth seems to sometimes alternate between 20 and 80,
>>>> but sometimes it stays fixed at 20. Hence, I was wondering what really
>>>> happens.
>>>
>>>
>>>
>>> For one thing, if the hardware detects interference on secondary
>>> channels,
>>> then it may decide to transmit a 20Mhz encoding.
>>>
>>> The rate-ctrl algs in the firmware differ from release to release, but
>>> the general goal is to transmit with the fastest encoding rate.
>>>
>>> Thanks,
>>> Ben
>>>
>>>>
>>>> Thanks,
>>>> Gaurang.
>>>>
>>>> _______________________________________________
>>>> ath10k mailing list
>>>> ath10k@lists.infradead.org
>>>> http://lists.infradead.org/mailman/listinfo/ath10k
>>>>
>>>
>>> --
>>> Ben Greear <greearb@candelatech.com>
>>> Candela Technologies Inc  http://www.candelatech.com
>>
>>
>> _______________________________________________
>> ath10k mailing list
>> ath10k@lists.infradead.org
>> http://lists.infradead.org/mailman/listinfo/ath10k
>>
>
>
> --
> Ben Greear <greearb@candelatech.com>
> Candela Technologies Inc  http://www.candelatech.com
>

_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

Re: Dynamic Bandwidth in hw rate control

[Ben Greear]

On 06/16/2016 03:44 PM, Gaurang Ramesh Naik wrote:
> Thanks Ben and Dave for your replies. So what I understand is that
> dynamic bandwidth management is a part of the rate control algorithm
> (which is implemented in firmware) and hence I have not much control
> over how that works.

With stock firmware you have very little control, though newer stock
firmware has some tweaks to rate-ctrl that the driver does not yet
support.

With my firmware and driver, you can at least disable any particular rate.
So, you could disable all 40Mhz rates, for instance.  I'm not sure how
useful that would be...

> I am assuming your firmware is the one that can be downloaded from
> http://www.candelatech.com/ath10k.php Is your kernel a modified one
> too? When you say I can see the tx rate for each frame, is it through
> some logs? I could try to get some info by using that. Thanks again.

Please read that page, it has links to kernel source.

tx-rate will be reported up the stack in a normal manner (ie, somewhat like
ath9k does it).  You could instrument the kernel to save a histogram of
tx-rates or something similar, perhaps.  'iw dev wlan0 station dump' and
similar will now show expected tx rate, but that is an average or maybe
the last tx rate reported, so it does not give a detailed view of
what is happening.

You can also sniff the air and see the rates in that manner (on any
firmware, including stock).

> Slightly unrelated, but is there a way I can play around with PIFS
> value? I could not locate it in the ath10k source, so I am assuming it
> is implemented in the firmware. I am trying to study some fairness
> issues when legacy devices operate in the secondary channels.

I don't have time to dig into this right now, but if you don't
see anything in the driver related to this, then probably it requires
firmware changes.

Thanks,
Ben


>
> On Thu, Jun 16, 2016 at 4:54 PM, Ben Greear <greearb@candelatech.com> wrote:
>> On 06/16/2016 12:57 PM, Gaurang Ramesh Naik wrote:
>>>
>>> Hi Ben,
>>>
>>> Thanks for your reply.
>>>
>>> I wanted to know if the dynamic bandwidth option implemented in ath10k
>>> is the same as that mentioned in the standard, or does it deviate? I
>>> tried to look for any documentation on ath10k dynamic bandwidth, but
>>> could not find any.
>>>
>>> If I am not wrong, on the secondary channel, the sender senses the
>>> channel for PIFS interval of time immediately preceding the start of
>>> the TXOP (9.19.2.8 in the standard). In the static mode, if the
>>> secondary channel is sensed busy, the sender waits until all channels
>>> become idle; but in the dynamic mode, the sender sends on the primary
>>> and (adjacent) secondary channel that are not busy.
>>>
>>> However, the documentation for dynamic bandwidth in ath10k/mac.c seems
>>> to suggest that the transmission retries are carried over narrower
>>> bandwidth in dynamic mode. Does this mean the same as what is given in
>>> the standard? Or is the dynamic mode implemented differently? Any help
>>> is appreciated.
>>
>>
>> I don't know exactly how this works.  At least the sensing and stuff is
>> probably
>> baked into the hardware.  What I notice is that the firmware passes a series
>> of rates to the hardware to use for each packet, and it includes different
>> bandwidth rates.
>>
>> If you used my firmware and my kernel, you can see the (successful) transmit
>> rate for each frame.  Perhaps you could discover some info about how it
>> works
>> in that manner if you could reliably add noise on an adjacent channel.
>>
>> Upstream firmware can see tx rates using pktlog I think, but I have
>> never tried that.
>>
>> Thanks,
>> Ben
>>
>>
>>>
>>> Thanks,
>>> Gaurang.
>>>
>>> On Thu, Jun 9, 2016 at 1:23 AM, Ben Greear <greearb@candelatech.com>
>>> wrote:
>>>>
>>>>
>>>>
>>>> On 06/08/2016 08:10 PM, Gaurang Ramesh Naik wrote:
>>>>>
>>>>>
>>>>> Hi,
>>>>>
>>>>> I needed a small clarification related to dynamic bandwidth setup. The
>>>>> documentation in wmi.h says " When enabled HW rate control tries
>>>>> different bandwidths when retransmitting frames."
>>>>>
>>>>> Does this mean that each packet is first always tried over VHT80 and
>>>>> if that fails it goes to HT40 and then to HT20? I tried this with one
>>>>> AP-Sta operating in VHT80 and one AP operating in its first secondary
>>>>> channel. When I check the rx rate info using "iw dev wlanx station
>>>>> dump", the bandwidth seems to sometimes alternate between 20 and 80,
>>>>> but sometimes it stays fixed at 20. Hence, I was wondering what really
>>>>> happens.
>>>>
>>>>
>>>>
>>>> For one thing, if the hardware detects interference on secondary
>>>> channels,
>>>> then it may decide to transmit a 20Mhz encoding.
>>>>
>>>> The rate-ctrl algs in the firmware differ from release to release, but
>>>> the general goal is to transmit with the fastest encoding rate.
>>>>
>>>> Thanks,
>>>> Ben
>>>>
>>>>>
>>>>> Thanks,
>>>>> Gaurang.
>>>>>
>>>>> _______________________________________________
>>>>> ath10k mailing list
>>>>> ath10k@lists.infradead.org
>>>>> http://lists.infradead.org/mailman/listinfo/ath10k
>>>>>
>>>>
>>>> --
>>>> Ben Greear <greearb@candelatech.com>
>>>> Candela Technologies Inc  http://www.candelatech.com
>>>
>>>
>>> _______________________________________________
>>> ath10k mailing list
>>> ath10k@lists.infradead.org
>>> http://lists.infradead.org/mailman/listinfo/ath10k
>>>
>>
>>
>> --
>> Ben Greear <greearb@candelatech.com>
>> Candela Technologies Inc  http://www.candelatech.com
>>
>


-- 
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc  http://www.candelatech.com


_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k