Collocating AP and BH in the same band

We are about to start our Canopy network.
We have to collocate an AP and a BH.
We plan to synchronize them by means of a timing cable.

We cannot use a different band (2.4 is crowded and 5.2 and 5.7 are not allowed here in Italy) and we have to locate the AP and BH on the same tower.

So, according to the Canopy User Guide, to avoid any interferences we should set the same Downlink percentage in both the AP and BH. But in this way we will under-utilize the BH bandwidth, since in the BH the traffic flows in the opposite direction of the AP.

If we use frequencies in the opposite sides of the 5.4 band (about 200MHz of separation), is it sufficient to avoid any interferences?

If you put the six AP’s on 3 frequencies on the quietest end of the spectrum, and put the BH at the other end you should be OK. Use the same frequency for opposing AP’s

Try to get the BH Master physically as far from the AP’s as possible

Use reflectors on the BH’s.

Ciao Gianni,

We do this without any problem. There is no need to have the BH and APs with the same download:upload percentages though it does make sense since at the end of it all the BH is what should be carrying the traffic that the subscribers will be allowed.

As Jerry explained, keep the BH as far away (both frequency wise and distance wise) from the APs and you should be fine.

Where are you based in Italy?

Am i correct is saying that you can not take a sync from a BH-Master. However if you take a Sync from a BH-S to feed an AP then you will be 180’ out of phase with the AP…

I never get away with sync problems…

Good question - let’s play stump the tech.

Who wans to email tech support with that one?

Im a little curious about frame structure differencies between those used by AP and BH. Basically it consists of uplink, guard band and downlink.
Why Motorola says BH should be placed far away from APs if operating in same freq band even if synchronization is used?
If we set the downlink ratio for APs and BHs to match the “Uplink Rcv SQ Start” value, it should not interfere if it is close to each other… :roll:
Is the frame length or smth else different?

Frame structure is different because there is no multipoint control on the BH.

Yeah, but in the means of downlink/uplink separation it should be the same. The basic frame is still 2.5 ms. AP’s uplink consists of many “uplinks” from different SMs, but its uplink anyway (AP Rx).
So if the frame length is the same and we match the guard time position on BH frame and AP frame, what can be interfering??
For my understanding, ‘multipoint control’ is just request/reservation process, which is included in basic frame anyway and does not effect its basic structure Tx/guard/Rx. :roll:

Motorola manuals does not say much about frames… So, Jerry, i would be more than happy if you explained how this ‘different frame structure’ effect the Tx Rx timing?

Actually, you would be correct. Assuming that the BH and the AP are set the same you would be in sync and would just need to deal with frequency spacing.

However, in a typical Canopy network your BH would have a larger uplink to the cluster and the AP’s would have higher downlink to the SM’s. By trying to force the AP’s and BH’s to the same uplink ratio you are hindering the potential of the system.

With that said, it sounds like you have a pretty good handle on it.

Thanks! I thought maybe there is something I missed.
BH or AP underutilization is not an issue in my case, because in this particular installation BH is not used for cluster uplink (it’s uplinked by fiber), but to connect a remote location, which cannot be reached by AP. So it’s ok, if the downlink ratio is somewhat the same.

I tried to use the frame calculator - the results show, that “Uplink Rcv SQ Start” values match perfectly, if I pick 60% for AP and 58% for BH. Of course, that also depends on APs control slot number and max range. Manual says, collocating APs you should care only about “Uplink Rcv SQ Start” value… What about other parameters? Of course, if APs with same max range, ctrl slots and downlink ratio are collocated, frames are the same. I’ve managed to match “Uplink Rcv SQ End”, “Downlink Rcv SQ End” values. But… what is the meaning of “Uplink Xmit SQ Start” i dont understand at all, and this parameter differs a lot (more than 1000).

Is the guard time always the same length? I got it 10 bit times (5.4GHz multipoint hw scheduler) in all calculations, that’s a bit strange, i was pretty sure, that “max range” parameter should influence guard band in some way…

Sorry for long story :wink: