Normally I am used channel 1, 6 or 11 (2412,2437,2462), My notes here if have for example 30 Ap as Caps. Is it permissible to re-use the frequency channel 1, 6 or 11 (2412,2437,2462) distribution for the same floor? without interference or confusion and you can see below the example of distribution "Caps", The diagram is simple, illustrative :
inside of building.jpg
You are at the holy grail of 2.4 GHZ wifi deployment. Many have tried and studied this deeply. There are numerous MUM presentations on this. But this is not a Mikrotik issue, it is the basics of 802.11 design and standard.
And those basics are good, bad and ugly. And they will frustrate and disappoint you (like anyone else in this wifi realm)
They are good because they deliver a wifi that always works, by slowing down until it works. There are slower speeds with lesser signal that goes further distance. And it goes very low, so the coverage is great. The whole design is based on a very polite rule, " only one talks at a time" (co-channel interference)
They are bad, because the usefull range for a decent speed is till -72 dB (see the sensitiviry of the hAP specs). The waiting for your turn goes down till a level of -96 dBm. So the usefull area around the AP is much smaller than the "wait your turn" area. Early on this was no problem, few AP's , low speed requirements. Now we have many AP's and need speed. The speed-span has risen from 1 to 11 Mbps to 6 to 300 Mbps. Those high speeds requiring higher signal levels (minimum -61 dBm) .
The ugly. Anyone in the lower signal range (-72dBm till -96dBm), can still communicate, but they talk slower, s l o w e r and s l o w e r . Worse yet they don't get the same share of time to talk , they get their turn to talk. And that 6 Mbps guy, takes 50 times more time to do its say. (If you still enable b mode, its 300 times slower). So the fast guy only gets 2% of the time to communicate, the slow one gets 98% of the time, giving them equal low speed.
The problem with 802.11 2.4 GHz is that the difference between usefull range and co-channel interference is too big.
In an open field, with a transmitter at 20 dBm, a receiver at 8 meter will typically receive -60 dBm.
In open field the signal drops 6 dB for double the distance.
4m= -54dBm
2m= -48 dBm
1m= -42dBm
0.5m= -36 dBm
16m= -66dBm
32m= -72dBm (the usefull speedy range)
64m= -78dBm
128m= -84dBm (stops usefull working)
256m= -90 dBm
512m= -96dBm (end of co-channel interference)
The ugliest part, are the cheaters that pump up the TX power above legal limits and add a strong antenna without compensating. 18 dBm more, 8 times the distance = 4 km disturbance.
For your channel reuse, which is a very good idea, please read this, and understand why walls might save the idea.:
http://divdyn.com/docs/2.4-GHz-is-Dead-v1.00.pdf