hi,
I need a help with Frequency reuse parameter. A have two e3000 back-to-back and im having some issue to find a good channel for the both.They have no smart anntena and only elevate sms. So i have a suspect of self interference and i read for this possibility.
My question is:
-i need only configure the both with same channel , one with back setor and the other front sector, gps sync actived and thats it? it will work?
Basically - yes. BUT, an understanding of HOW that works is key. The idea is that:
a) the TX portions of the frame are all syncronized at the same time, and then the AP's all go quiet and go into RX mode, and then each SM talks in it's schedule time slot. So, that helps each AP not interfer with the other AP's -- since one AP isn't trying to listen to it's client, at the exact instant another AP 3 feet away is transmitting. With GPS sync, that AP to AP interference is mitigated.
b) The SM's all take their turns, on schedule, so on each Sector, the clients should also not bump into each other... so SM to SM (on the same Sector) interfernce is eliminated.
BUT - here's the key - you want all the SM's to have at least -30 dB difference in the Sector it's connected to, vs the Sector pointed the other way. If you don't have good isolation, things can still self interfere. In other words, if SM#6 on Sector-East is transmitting, and if SM#14 on Sector-West is transmitting at the same time... if Sector-East can still hear the SM's which are connected to Sector-West too loudly... then things can still get screwed.
So - yes - turn on GPS, set one to Front and the other to Back... but it's also important to make sure that each SM can't hear both sectors too loudly (ideally -30db) and that each sector can't hear both groups of SM's (ideally -30db).
@ninedd wrote:Basically - yes. BUT, an understanding of HOW that works is key. The idea is that:
a) the TX portions of the frame are all syncronized at the same time, and then the AP's all go quiet and go into RX mode, and then each SM talks in it's schedule time slot. So, that helps each AP not interfer with the other AP's -- since one AP isn't trying to listen to it's client, at the exact instant another AP 3 feet away is transmitting. With GPS sync, that AP to AP interference is mitigated.
b) The SM's all take their turns, on schedule, so on each Sector, the clients should also not bump into each other... so SM to SM (on the same Sector) interfernce is eliminated.
BUT - here's the key - you want all the SM's to have at least -30 dB difference in the Sector it's connected to, vs the Sector pointed the other way. If you don't have good isolation, things can still self interfere. In other words, if SM#6 on Sector-East is transmitting, and if SM#14 on Sector-West is transmitting at the same time... if Sector-East can still hear the SM's which are connected to Sector-West too loudly... then things can still get screwed.
So - yes - turn on GPS, set one to Front and the other to Back... but it's also important to make sure that each SM can't hear both sectors too loudly (ideally -30db) and that each sector can't hear both groups of SM's (ideally -30db).
Hi ninedd,
First off, excellent explanation of said topic! I had a good grip on the parameters but your explanation solidified everything in my mind and confirmed I fully understood. Just want to confirm, when you say the SM cannot "hear" both the sector it is attached to and the sector behind, you can see if that is happening by looking on the SMs "Monitor"->"Wireless Tap" and you would see the offending sector(the back one) and make sure its at least -30db difference from the setor it is attached to?
ninedd,
Thank you for perfect explanation!
Just 2 cents from my side:
30dB(reason to have Antenna F/B ratio 30 dB) difference is required for SMs to be able to operate on MCS15(802.11n) and 33-34dB(reason for ePMP3000 Sector Antenna to have 35 dB F/B) is required for 802.11ac SMs to be able to operate on MCS9 DS.
Thank you.
Thankyou for the great Answer,
Im working with cambium about 2 years now and every day i learn a new thing.
I dont know if i got you. Like i said before, my problem this time is too much noise. I already jump for many channels and dint work very well.
I see in spectrum analyser tool that the other epmp 3000 back-to-back in this tower is interfer.
How can in know that the reuse frequency works? Its necessary really put the same frequency ?
Some informations: We use 40mhz, and 30db of transmitt power on both AP's.
@Vitorhugomb wrote:Its necessary really put the same frequency ?
,,,when you say the SM cannot "hear" both the sector it is attached to and the sector behind, you can see if that is happening by looking on the SMs "Monitor"->"Wireless Tap" and you would see the offending sector(the back one) and make sure its at least -30db difference from the setor it is attached to?
If I understand your question - no, it's not necessary to do frequency reuse. We have places which are remote and quiet, and we use 4 different frequencies for the 4 different sectors. If you've got the channel space, that's 100% fine too. Usually (in today's noisy world) that's often not possible - so frequency re-use allows using the best channel for the AP, and use it on both back-to-back sectors (as long as everything is right).
And if I understand the other question, yes - if you look at the 'Montor Wireless" and if you see both Sectors, and if they are -60 (the one you're facing) and -70 (the backside of the other) then that's not going to work too well for frequency reuse. If that's the case, the SM will get ''interference'' from the other Sector when it's broadcasting to it's clients. Likewise, that would mean that the 'back' sector is hearing your SM too. So, while it's trying to listen for it's clients, it'll hear your 'Front' SM talking to the 'Front' Sector, and so the 'back' sector' will be interfered with by the 'front' SM too.
So - for Front/Back frequency reuse, the better the isolation the better. You don't ''absolutely need'' -30db difference, but that's where the highest modulation and best performance comes in.
thanks alot,
I will put this in pratice and see how works.
One thing that i've notice is the large diference betwen using Smart Antenna vs not using, its unreal . Even epmp 3000 that i particuly consider the best ap that i work with it, do not work very well in some cenarios of interferency.
We have about 8 towers with epmp's. In some cases three epmp 3000 side by side but all with beanforming and works perfectly well.
But again thanks for the help!
Do not forget proper horizontal separation between ap antennas. The idea is that you must place each antenna in the back null of the opposite antenna. The 30db f/b ratio is only good if you are in that null. So depending on the actual sector antenna used you may have to have 3m (10ft) of horizontal separation. We are using KP performance sectors with the radio box on the back and found that 2m horizontal was required to remove the interference from the opposite ap.
Other than the above, you have some really good answers here.
@Douglas Generous wrote:Do not forget proper horizontal separation between ap antennas. The idea is that you must place each antenna in the back null of the opposite antenna. The 30db f/b ratio is only good if you are in that null. So depending on the actual sector antenna used you may have to have 3m (10ft) of horizontal separation. We are using KP performance sectors with the radio box on the back and found that 2m horizontal was required to remove the interference from the opposite ap.
Other than the above, you have some really good answers here.
Hi, so how did you find out that the opposite ap was giving interfernece?
First sorry for the very late reply,
Radio engineering has a few rules/axioms that help determine the most likely issue location. As you gain experience and see different situations you will learn what the most likely issue point is, especially if youve been bitten by it before. An example of this is when I deployed our first epmp 5Ghz equipment using kp sectors on a single pole back to back. The radios were set to the same frequency and one was back the other was front, syncd to a packetflux box. They worked but strange things with deregistration and such was causing havoc. Went through everything, even called KP to make sure it wasnt defective antennas. The their engineer asked me about my null field separation, which I explained was twice the mounting distance plus 3 inches, zero verticle. Then I asked what the resonant distance to the null behind the antenna was since its not usually on a data sheet nor is usually needed. Turns out that its 3ft. This is important when mounting antennas back to back as metal just is a reflector but another resonant object can both amplify and distort the signal. By moving the antennas out 3ft each our problem was eliminated. On all antennas some power is leaked behind the antenna, even though the antennas have a 30dB f/b ratio.it was enough to cause issues.
On top of experience I also read a lot, I go through books at a rate that makes others question my sanity.
May your thirst for knowledge leave you parched.
No need to apologize at all! So you have a total of 6ft horizontal space between the back to back radios?
No, not necessarily… ideal distance between antennas depends on radio power output, quality and specs of the antennas, tower composition, proximity of clients and their EIRP, etc. As a rule of thumb however, having more physical separation between antennas in a reuse situation is… how Martha Stewart would say… “It’s a good thing”.
For that setup I had 6ft between the backs of the antennas, but now we consider it a minimum. There is a lot of little things to consider such as the frequency, the antennas f/b ratio, allowable eirp, actual eirp. Tower type and physical characteristics even the cables and their leakage rates must be considered. Even other antennas on the tower all play a role in how this works to some degree.
One broadcast tower we are on it 2m wide, this is enough to simply direct mount to, but its also 180m tall and acts as a resonator (old vhf+uhf tv tower). This means we must be at least 1m away from the tower surface or it could reflect our back signal and cause troubles like its doing for another colocator whom direct mounted and blames our install as the issue.
