Over the past year or so, I have been getting some complaints from customers connected to our Epmp 2000 APs that are GPS sync'd. Every now and then, some of these tickets have me scratching my head.
I'll run an edetect that shows our opposite facing AP's SSID multiple times as an interferer, which are the SMs attached to that AP. So I'm assuming those SMs are causing interference to the opposite facing AP.
We are running 4.4.1.
GPS firmware is up to date.
Appropriate settings match for the GPS sync'd APs.
#1 Am I correct in this assumption?
#2 What fixes for this situation are there aside from changing the AP that the SMs are connected to?
Are you running the edetect from the AP? If so it will show the AP behind it as interference, but if you have gps setup correctly for back to back then it won't interfer.
Make sure one AP is set to front Sector, and the other to Back sector - also that they are both on the same frame size.
What do you have set on the AP for Subscriber Module Target Receive Level? Setting this too hot will make it so the back AP will hear the CPE's on the opposite side of the tower, which would cause interference. We typically have this set at -60 in our network, and we are using cambium antennas for the F/B ratio.
Yes, running edetect from the AP in question shows Wireless MACs of SMs attached to the opposite facing AP that is sync'd.
One is set to front, the other to back.
Same frame size.
We use -60 as the target receive as well.
I see my options as being either disabling GPS sync and running different frequencies
or
maybe turning down the interfering SMs tx power down. They are all connected in the -40's and -50's on auto-adjust, so there is plenty of breathing room to adjust the tx down manually. This is assuming that this is actually causing the issues.
If I'm just an idiot looking at this all wrong, do let me know
When you set the Subscriber Module Target Receive Level to -60, that automatically adjusts the CPE's transmit power so that should be covered.
Are you sure theres not outside interference thats happening here? Do customers from both AP's have issues? What MCS rates are you seeing at the access point? Can you move one AP to another channel? I don't think GPS is causing you any issues, but there could be outside interference in the area thats causing intermittent issues.
When you scan using eDetect it only scans the exact channel and size you're in, so if theres a overlapping interference from someone using a 40Mhz or 10Mhz channel it won't see it there.
If I'm just an idiot looking at this all wrong, do let me know
Not an idiot by no means, but a little confused. My best suggestion is to read and digest the attached pdf. After doing so, come back with any questions and I am sure we can help you.
I can tell you this, if you have towers close to each other, using the same freqs between towers, disabling GPS is NOT the answer.
I appreciate you guys taking time out of your day to help me out here.
I understand that particular shortcoming of edetect. I have read the pdf before, and I believe I have at least a decent understanding of how to operate GPS sync and channel reuse. I'll go over it again, maybe I missed something there. This tower actually is pretty close to another tower site that we have that also has the same equipment deployed. If I did disable GPS sync and channel reuse, I would not run the same broadcast channels on my APs. I'm pretty terrible about being clear through text.
AP-west is sync'd with AP-east.
The appropriate radio configs are matched across sync'd APs: channel width, DL/UL ratio, frame size, frequency carrier, channel width, target rx level, sync source(GPS). AP-west is back, AP-east is front.
AP-west is picking up several interferers which are SMs connected to AP-east, also AP-east itself. I know that although AP-east shows as an interferer, it isn't actually causing any interference to AP-west because their transmit and receive timing are synchronized.
Correct me if I am wrong, but using GPS sync, when AP-west is receiving, AP-east's SMs are transmitting in that moment, right? So the SMs on AP-east could be interfering with AP-west.
I changed the target rx level for both sides to -65 since all of the SMs had -55 or better on DL and UL. As far as edetect can show me, most of the interfering SMs are not any more.
I believe that both west and east AP's will transmit and recieve at the same time. So if East can see some customers that are attached to west with signals that are too hot that would be causing interference. Same with if West can see customers that are attached to East. You could try lowering the power on those even more that what the auto-adjust is doing as long as you keep the modulation levels high.
We've run into that issue where we have a lot of customers very close to the tower, usually in those situations we change one AP to another channel, but we always leave sync so that it doesn't interfere with our other towers in the area.
Yeah, even if you decide that you’re not going to reuse frequencies, there is no reason not to sync. You should always use GPS Sync, even if you use 4 different frequencies on 4 sectors.
> You should always use GPS Sync, even if you use 4 different frequencies on 4 sectors.
I think this is/was probably true with Cambium 450 and old Motrola gear ( don't recall Canopy GPS ever outright failing in the 13'ish years we have been using it) but remember that when you are using GPS sync and sync stops working you now have an outage because the AP stops transmitting. So given ePMP's various problems with losing GPS sync, thus dropping all connected customers, I don't think I would recommend using ePMP GPS Sync unless you actually need GPS Sync.
> You should always use GPS Sync, even if you use 4 different frequencies on 4 sectors.
...given ePMP's various problems with losing GPS sync, thus dropping all connected customers, I don't think I would recommend using ePMP GPS Sync unless you actually need GPS Sync.
YMMV. For what it's worth, I have never had a single incident of GPS Sync being lost on ePMP (other than early 'beta' firmware). I know other's have reported that, so it must happen for some people in some situations, but I've never seen GPS problems on our live network.
So for my 2c, I woudn't recommend avoiding GPS because you're scared of maybe it might fail.
" I have never had a single incident of GPS Sync being lost on ePMP (other than early 'beta' firmware). "
Really ? Is 3.5.2 RC15 "early 'beta' firmware" ? The best part is you was running GPS on an epmp AP that you didn't need to be running it on and just as I point out, it caused repeated unnecessary outages.
I went looking because GPS problems are so bad on ePMP I found it completely unbelievable that anyone using ePMP GPS for any length of time has not had GPS fail...
"In any event, last week the GPS lost link (which disconnected the clients of course) and we rebooted and no change. I upgraded the firmware to 3.5.2 RC15 and when that booted up, the GPS was back - with SNR's in the 40's
It's gone away a few times since, and no - I haven't yet tried disconnecting the external puck. And no - I haven't yet tried replacing the AP with another unit. it's quite a climb and I'm trying to get a feeling for what the likely culprit this is before I start changing hardware.
This is the GPS Monitor right now - with 0 tracked satellites."
OH, OK. I guess I had forgotten about that. As I remember, I upgraded firmware, and I haven't had GPS problems since. The current uptime on that AP is 17 days (since the last power outage).
I will also add that we did have issues on 4.4.1 while testing 3000L at the shop, but not on anything we deployed with clients connected. We've been on 4.4.2 for 12 days with 3000L's in the field, and no GPS issues on 4.4.2 so far. I guess 12 days isn't a long test, but that's how long we've had 4.4.2 running - so that's all I can say.
Anyway, certainly for us, GPS hasn't been something we've had to babysit or worry about. I actually had forgotten that we had had those GPS problems 18 months ago.
Gregorius, it sounds to me like the problem was simply insufficient front to back ratio in the sector antennas. What kind of sector antennas were you using? Lowering the target power does reduce the unwanted signals, of course, from SMs that are very close and thus overwhelming the F:B of the opposite sector. RF Armor sometimes helps too.
