On 08/13/2015 03:40 PM, Chuck Enfield wrote: > Just to be clear, we don’t have to do these things to make wireless > work. It makes it work better. But it is sometimes necessary to make > wireless work acceptably in the most challenging environments. That > said, left to defaults Aruba’s ARM also adjusts 2.4 GHz Tx power way > down. So far down, in fact, that coverage gaps show up. It also tends > to keep power higher on busy APs and lower on less busy ones. In some > cases the power on the 2.4 radio will be the same as the power on the 5 > GHz radio. These characteristics forces us to configure a range of > acceptable power levels for ARM to choose from. Once you’re doing that, > why not select the optimum power levels? > > While I’m no expert on Cisco wireless, I have assisted some departments > with problems on their Cisco infrastructure. Based on that limited > experience, I have far less confidence in RRM than you seem to. >
The main issue is the defaults for Aruba are for coverage networks, not high density (30-40 ft) or very high density (30 Ft or less). You need to adjust the coverage index min/ideal for high density deployments so ARM will power down to Min TX powers without shutting down the 2.4 GHz radios due to CCI (even with edge detection for the APs the default will still end up with only a few 2.4 GHz APs). If you set Tx Min/Max to a 6dBm range the APs can power to around double the cell size for coverage gaps if a radio is down. I also like to set the 5 GHz Tx Min/Max range at least 3dBm higher then 2.4 GHz because of the unattenuated propagation distance and better attenuated penetration. Which also helps duel band devices make better selections. However most devices have a fairly generous threshold on AP signal drop before they even try and probe for candidate APs to associate to. That's really where the controller based client stearing solutions come in to play with selective acks or the probes to get the client on the best AP regardless of what the client wants based on just Rx signal. -- -James ********** Participation and subscription information for this EDUCAUSE Constituent Group discussion list can be found at http://www.educause.edu/groups/.
