cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
3890
Views
10
Helpful
9
Replies

Access Point Height , RRM (Auto RF) and Warehouse

desaikeyur
Level 1
Level 1

Hello Experts,

I have a Warehouse situations where there is a high density Access Point deployment. The APs are 1242 with external antennas. The height at which these antennas are mounted is 30 feet.

The clients (Scanners) are complaining poor signal. I also see several Access Points powered down to level 8 due to high density.

My question is is it possible that due to Auto RF (RRM) that the APs have reduced the power such that by the time signal reaches the client it is very week.

What is the maximum height allowed on these 1240s in an indoor environment.

Has anybody seen something like this.

9 Replies 9

Leo Laohoo
Hall of Fame
Hall of Fame

Are you using autonomous or controller-based AP?

Have you done a walk around with a spectrum analyzer to find out if the channels are interfering with each other?

Aaron
Cisco Employee
Cisco Employee

This design is a concern. Note that RRM's power adjustment is based upon how loud the APs hear each other, not how loud the clients hear them. So it is quite possible that your APs hear each other very loud, and are reducing their power to the minimum, while not much RF energy is reaching your clients.

What antennas are you using? How are they oriented? It would be nice if your antennas were somewhat directional (patches, perhaps), and were aimed down towards the clients.

You really should survey the coverage area and get a sense for how many APs your clients can hear, and at what signal levels.

It is possible that RRM may not be optimizing your power levels, given your physical environment. If the power levels are set too low (compared to what the clients on the floor are hearing), you can adjust the TX power control threshold setting, for example:

config advanced 802.11b tx-power-control-threshold -55

The default is -65 or -70 dBm (depending on version); for dense indoor deployments, it may make sense to lower this value, to say -73; however, in a case like yours, it may make sense to raise this value.

The antennas are omnidirectional 2db antennas. And the warehouse is used for scanners.

The scanners have the following specs

Transmit Power 32mW max (+15dBm), all versions

Frequency Range 2.400 - 2.4835 GHz, all versions

Channels 1 to 11 for USA, Canada

1-13 for EU countries

RX Sensitivity (8% FER, 1024 bytes packet) =

-89dBm @ 1 Mbps, -87dBm @ 11 Mbps,

-83dBm @ 6 Mbps, -72dBm @ 54 Mbps

I am going to try your recommendation and check the threshold and see if that makes any difference. I am also concerned about the co channel as scanner use g only.

> The antennas are omnidirectional 2db antennas.

OK, that may not be so great for this application. How are these

omnis oriented? Bear in mind that the gain of an omni is perpendicular

to the antenna, and that there is a null in the direction of the antenna.

Let me try to explain this with some ASCII art (hope this works out).

Let's look at a 2dBi "rubber duckie" omni such as the AIR-ANT4941.

(See http://www.cisco.com/en/US/prod/collateral/wireless/ps7183/ps469/product_data_sheet09186a008008883b.html

for more details on the antenna's actual radiation pattern.)

Let's say that the antenna is pointing straight down from a

ceiling-mounted AP:

__________________ __________________

| | |

| AP | | neighbor AP

| | |

__________________ __________________

||

||

+2dB <||> +2dB

||\

|| \

|| \

|| v

-20dB -10dB

||

client

So, the signal as perceived by the client right underneath this AP is

attenuated by 20dB, and the signal as perceived by another AP

mounted on the same ceiling will have 2dB of gain.

So - the APs hear each other extra loud, and the clients on the

floor below hear the APs extra quiet.

Now let's take that 2dBi rubber duckie and bend it by

90 degrees:

__________________ __________________

| | |

| AP | | neighbor AP

| | |

__________________ __________________

|| +2dB

|| ^

»»»»»»»»»> -20dB

v

+2dB

||

client

So now, the radio energy has 2dB of gain directly below the AP, and the

-20dB null in the direction of the antenna is pointed at the neighbor AP.

So you have more of your 802.11 energy aimed at the clients (where you

want it), and less aimed at your neighbor APs (where you don't want it.)

After changing your antenna orientation, you would want to resurvey,

and then retune your RRM tx power threshold as needed.

Hth,

Aaron

Aaron
Cisco Employee
Cisco Employee

Looks like Netpro horked my carefully wrought ASCII art ... so see the attachment.

Thanks Aaron, I am leaning towards changing the RRM settings. But here is some more information that can help you assist me better.

Changing Antenna is not an option. Also I believe the null under one antenna should be covered by another AP in the neighborhood. Unless the AP height of 28ft - 30ft,causes the signal to drop drastically by the time it reaches the client.

Thanks - from the picture in your doc file, it looks like the AP / antenna installation are reasonable.

The next step is to get down on the ground, and see how the coverage looks there. I'd start with a trusted site survey tool (I use a CB21AG with the Cisco Aironet Site Survey Utility, though a pro might use Airmagnet.)

If the coverage looks good in the locations where the client is suffering from a problem, then the next step is to do a wireless sniff.

It would also be a useful test to try a different wireless client (e.g. a Windows XP machine with a CB21AG), configure it just like the handheld, and see whether or not it also sees a problem.

I'm currently working in an environment similar to what you say, with the same access points, the same omnidirectional antennas and the same distribution and location of antennas (30ft).

The problem I have is interference with other wireless networks, but I think that could be solved, at least in part, by varying the parameter TX power treshold according you propose. I think that the APs can hear each other very well so their antennas are located quite high, but clients, especially when there are other wireless networks, do not hear so well the APs from the floor.

Do you think that if I change this parameter I can solve the problem?

I tested with the parameter TX power treshold in -65 and -70. Should I therefore upload it to -55 or reduce it to -80?

Thanks

The normal assumption that RRM makes is that the APs can hear each other about as loud as the clients can hear the APs. On that assumption, one will normally want to set the TX power threshold to -70dBm or so. (A little less, e.g. -73dBm, in a very dense deployment; a little more, e.g. -65dBm, in a sparse deployment.)

If that assumption is wrong, and the APs can hear each other much louder than the clients can hear the APs - that's where you'd want to *raise* the TPC threshold. E.g. to -60 or even -55.

As always, I would want to validate the results via some kind of site survey on the ground.

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: