Design Surveys and RRM

[jamesconsterdine]

Hi all

We've had quite a few problems with RRM, in particular when we've previously engaged contractors to do AP Placement designs they've failed to consider RRM. So they do their survey with test APs on static power levels, but because of RRM those power levels are not reached when the service goes live. 

The best example is buildings where they designed the APs to go in corridors, expecting them to run at 12dBm, but because the APs all have line of sight of each other they turn down to power level 6 on the b/g/n radio (so around 3dBm generally) and coverage is insufficient. Even setting the TPC power threshold to the most aggressive -55dBm they still see too many neighbours and end up on power level 6. We don't want to turn RRM off and use static levels as it is a useful feature for many reasons.

As a result, we are now asking the design contractors to consider RRM, and make sure that the power levels they use in their survey are close to the likely power levels that will be negotiated once RRM is in place. 

The response we are getting is that it's "impossible to predict what RRM will do". I disagree with this statement, the TPC algorithm is not that complicated (essentially tuning so the 3rd strongest neighbour is at the TPC threshold level), so whilst you can't predict with 100% accuracy, I do believe it should be possible to consider with the right amount of effort and time being put in. 

Am I being unreasonable to think that any company designing AP placements for a Cisco Enterprise solution should have the knowledge to consider this feature?

Do Cisco provide design companies with any tools to help them predict what RRM will do? 

Any thoughts welcome.

Thanks

James

[Leo Laohoo]

The response we are getting is that it's "impossible to predict what RRM will do". I disagree with this statement, the TPC algorithm is not that complicated (essentially tuning so the 3rd strongest neighbour is at the TPC threshold level), so whilst you can't predict with 100% accuracy, I do believe it should be possible to consider with the right amount of effort and time being put in. 

I agree that it's difficult to predict what RRM will do.  The reason is the surrounding WiFi changes.  Take note, I'm not yet talking about non-WiFi interference (like BT or microwave ovens).  

I mean consider if the site is co-located with another business.  During the site survey, that business only had a cheap WiFi solution.  Plus 6 months and that business upgraded their wireless.  At least with Cisco's WLC, you can enable TPC/DCA and the controllers will take appropriate actions to reduce co-channel interferrenc.

[jmeachum]

James,

Assuming the survey was done properly, you should be able to set the minimum TPC power level to be equal to the survey power used.  The survey should also have knowledge about the devices you are using/planning on using on your network.  This will determine what the maximum TPC level should be set to, which is the TX power of the weakest transmitting device (The survey TX power should be less than the max TX power of your weakest transmitting device.).

So, if the weakest transmitting device is 50mW and the survey was done at 10mW then you will want to set your TPC settings to a max of 17 and min of 11.  This will guarantee that the coverage required based on the survey provided and allow APs to increase TX power when you have coverage hole detections or the TPC power threshold is not met.

The ideal situation is that you would not do hallway deployments for APs (another topic), but if you do then the TPC settings must be tweaked to match the design.

John Meachum