Thanks Jason and Jeff,

It is really confusing, especially with conflicting information out there,
like with the hysteresis of being 3 dBm both ways or 3 dBm one and 6dBm the
other.

Also after calculating, how would you know to increase or decrease the
power?

The following is how I understand it after some thinking. Some of the
figures are just base on real world experience. Say you do a voice survey
and place your APs to provide coverage of -67 dBm, 20% overlap and power
level of 6-25 mW. The following are the assumptions:
- Cisco formula as mentioned above is correct
- TPC only decreases power
- Hysteresis of 3 dBm
- There's another APs see each other rather than by transmitting neighbour
message at full power

Starts off at level 1 / 100mW. I chose the 3rd highest RSSI to be -50 dBm
as that would happen in a voice/high density survey.

20 - (-67 - (-50)) = 20 - (-17) = 37           Hysteresis met, decrease
from level 1 / 20dB to level 2 / 17dB
20 - (-67 - (-60)) = 20 - (-7) =  27            Hysteresis met, decrease
from level 2 / 17dB to level 3 / 14dB
20 - (-67 - (-70)) = 20 - 3 =  17               Hysteresis met, decrease
from level 3 / 14dB to level 4 / 11dB
20 - (-67 - (-80)) = 20 - 13 =  7               Hysteresis met, decrease
from level 4 / 11dB to level 5 / 9dB
20 - (-67 - (-85)) = 20 - 18 =  2               Hysteresis not met. Stay at
level 5 / 9dB

At level 5, it is close to the figures of some deployments out there.

Does this calculation make sense? Did I miss anything? If it correct,
great! Just need to confirm the assumptions above then.

Thanks
J Chew


On 12 September 2013 01:39, Jeff Rensink <[email protected]> wrote:

> This is a frustrating topic indeed.
>
> TPC will do both power decreases and increases as the environment changes.
>  But the formula that the document references has a flaw that I haven't
> totally reconciled.
>
> Tx_Max for given AP + (Tx power control thresh – RSSI of 3rd highest
> neighbor above the threshold)
>
> Tx_Max for a given AP should be a static value, correct?  It's supposedly
> the max possible power that the radio can transmit at for its chosen
> channel.  The Tx power control threshold is also a static value that is
> configurable.  So the assumption is that the RSSI of the 3rd loudest
> neighbor changes along with the current AP power level.
>
> If I understand things correctly, the RSSI of the 3rd loudest neighbor is
> determined by AP neighbor messages.  According to documentation (and I
> believe I've tested this myself), neighbor messages are always sent out at
> the highest power level and lowest data rate.  If that is indeed true, then
> the RSSI of the third loudest neighbor is yet another static value
> (assuming the environment isn't changing).
>
> So if none of the variables in the equation are actually variable, the
> results of the equation are always the same.  That would mean either the
> power level would always settle into the max power or the minimum power.
>  That's not true, so something seems off in the equation or the description
> of the variables in the equation.
>
> My guess is the formula is closer to (Max_AP_Power - Current_AP_Power) -
> (3rd_loudest_neighbor - TPC_Threshold)
>
> A positive result would result in a power increase.  A negative result
> would result in a power decrease based on the 3 dBm hysteresis.  According
> to Jerome Henry, it's 3 dBm in both directions in the 7.0.116.0 code.  This
> assumes that the 3rd loudest neighbor measurement is pretty much always the
> same since neighbor messages are always sent as max power.
>
> So let's say we start at full power on the 2.4 GHz band with a 3rd loudest
> neighbor of -60 and threshold of -67.
>
> (20-20)-(-60-(-67)) = 0-7 = -7
>
> -7 is lower than -3, so we drop the power down a level
>
> (20-17)-(-60-(-67)) = 3 - 7 = -4
>
> -4 is lower than -3, so we drop the power once more.
>
> (20-14)-(-60-(-67)) = 6 - 7 = -1
>
> -1 is not lower than -3, so it doesn't change.
>
> This is a total guess, but it makes more sense to me than the formula in
> the documentation.  But in the real world, the higher the TPC threshold,
> the higher the power levels will be.  For instance, a threshold of -60
> would result in higher power levels than a threshold of -70.  So if you are
> seeing your APs in a particular location settle in at too high or too low
> of power levels on average, tweak the threshold as needed.  Every increment
> of 3 should result in powers going up/down one level on average.
>
> Jeff Rensink - CCIE #24834 (Wireless, R&S)
> Senior Technical Instructor - IPexpert
>
>
> On Wed, Sep 11, 2013 at 9:14 AM, Jason Boyers <[email protected]> wrote:
>
>> This is a confusing algorithm.  Basically, think of it this way:
>>
>> 1) Three neighbors (at least) hear an APs radio at or stronger than the
>> TPC Tx mac value configured under the WLC TPC settings for that PHY
>> (802.11a/n or 802.11/b/g/n)
>> 2) If #1 is true, then that formula provided is used.
>> - The "Tx_Max for given AP" is the maximum transmit power for that
>> channel on that particular radio.  So, for a 2.4GHz radio, that would be
>> 100mW (20 dBm), since that is the maximum for 802.11b.  5GHz is more
>> confusing, as it various based on channel and model of AP.
>> - The "Tx power control thresh" is the value that is entered in the WLC
>> - Note that there is an "hysteresis" value of 3dB for decreasing and 6dB
>> (or it was at one point - someone correct me for the current values) for
>> increasing the power levels.  In other words, the value that is given at
>> the end of the formula has to be at least that dB difference from the
>> current value in order for a change to occur.
>> 3) The AP's radio that is heard by the other three APs is the one that is
>> decreased (or it can be increased) if the hysteresis value is exceeded.
>> 4) When the formula is run again, it is presumed that the three neighbors
>> that are hearing the APs signal are hearing a weaker signal.
>> 5) Keep in mind that the algorithm is always from the perspective of an
>> AP being heard by 3 or more neighbors, not an AP hearing 3 or more
>> neighbors.
>>
>> Hope this helps.
>>
>> Jason Boyers, CCIE #26024 (Wireless)
>> Blog: netboyers.wordpress.com
>>
>>
>> On Wed, Sep 11, 2013 at 9:36 AM, Jeen Sern Chew <[email protected]>wrote:
>>
>>> Hey guys,
>>>
>>> I have a TPC question.
>>>
>>> There are many conflicting information from different Cisco docs. Some
>>> say TPC increase and decrease power, some say TPC only decreases power
>>> while CHD increases power. Does TPC do both or only increase?
>>>
>>> Also Cisco gives the TPC formula of:
>>> Tx_Max for given AP + (Tx power control thresh – RSSI of 3rd highest
>>> neighbor above the threshold).
>>>
>>> Is the Tx_Max is the Tx_Max of the specific AP? or Is it the Tx_Max
>>> configured on under TPC in WLC?
>>>
>>> Also, when the calculation is done, I am assuming the power
>>> increase/decrease occurs on the neighbour with the third highest RSSI?
>>>
>>> Thanks
>>>
>>> Regards,
>>> J Chew
>>>
>>> _______________________________________________
>>> For more information regarding industry leading CCIE Lab training,
>>> please visit www.ipexpert.com
>>>
>>> Are you a CCNP or CCIE and looking for a job? Check out
>>> www.PlatinumPlacement.com
>>>
>>>
>>
>> _______________________________________________
>> For more information regarding industry leading CCIE Lab training, please
>> visit www.ipexpert.com
>>
>> Are you a CCNP or CCIE and looking for a job? Check out
>> www.PlatinumPlacement.com
>>
>>
>
_______________________________________________
For more information regarding industry leading CCIE Lab training, please visit 
www.ipexpert.com

Are you a CCNP or CCIE and looking for a job? Check out 
www.PlatinumPlacement.com

Reply via email to