Can client devices reduce their tx power according to the RSL at the AP? Is this in any amendment or based on proprietary technology?
I'm trying to limit the impact that individual client devices have on the contention domain size in an high-density outdoor deployment.
If I set the tx power of the AP to its min of 4 dBm in order to minimize cell size in open space, and space APs operating on the same channel to limit CCI (-82 dBm CCI boundary for OFDM), would clients continue to transmit at their max power (ie, 14 dBm worst client) if they are near an AP? In other words, is there a mechanism built-in to reduce the tx power of clients with good data rates near APs? I suppose this is partly mitigated by the slight reduction of power at higher mods, but would like to know if clients automatically reduce their tx power past their tx power per mod specs, or if there is no control over their tx power whatsoever.
802.11h addresses Transmit Power Control.
The Wi-Fi Alliance's search tool lists 5,330 devices that support it.
For WLAN's, the property you are talking about is either the older RSSI or the newer RCPI value (802.11k).
Neither value is especially accurate. Although I don't have any personal experience with it, the RCPI number should be better.
For what it's worth, some lower cost devices do not change their output levels based on the modulation . This kind of control requires better radios and/or more testing to get right. So to keep prices down they simply use the same (lower) level for all the rates.
Sad, but that's the real world.
Jerome Henry made an Ekahau presentation last week on Apple products and I-Phone iOS performance that you would probably find interesting:
In addition, CWNP has a webinar on the mysteries of Roaming coming up:
The iBwave seminar on that same page was also EXCELLENT and you'll probably find it useful in understanding your situation.
Thanks @Howard. Since most of the cheaper devices don't manage their Tx power, what is the impact to performance from having a Tx power mismatch where, instead of the AP, it is now the client who transmits at a higher power? Could clients unnecessarily blacklist SSIDs after sending multiple (5+) requests because the AP responds with too low a tx power?
I really enjoyed Jerome's presentation by the way; thanks for sharing.
I use iBwave... it's a great tool. Would love to see an integrated protocol analyser.
I would think that anytime you have mismatched power levels that you could have issues - I have really only experienced the AP's having greater power, so I can't say from experience the results of the opposite condition.
Generally speaking/ I would think that either the clients wouldn't find an AP to connect to, or that the client would always want to roam.
Some client devices have many configuration parameters for tuning such things as roaming, and others. However, that does not mean that the parameters are easily identified or modified by the end-user. Often only the third party installers or VARS are aware of the options. They will often make custom changes for their customers, especially in very large deployments.
You can bet that any company installing client devices at 10,000 stores will have a very close relationship with its manufacturer.
Take for example a DAS installation. Due to the peculiar (IMO) RF environment, a client tuned for a well tuned, strictly Wi-Fi installation, will likely not perform very well with DAS. You can bet the roaming thresholds (sticky or not etc.) will not be optimal. If you understood all of the tuning parameters however, and could access them, you could probably make the client perform as well as any other.
According to Jerome's presentation you could have clients blacklist the AP's, but I am betting that if you have a good micro-cell design that there will not be a problem. I doubt the client can keep a long enough list to track every probe request it has ever made. Closer AP's should always be fulfilling the needs of the clients, and it might never have 5 successive probes refused. Right now, this is only iOs devices anyway.
All in all, a very interesting subject.