hi,

can i use 2 different antennas with MIMO 2x2 card (RB711 / R52Hn) ???

how will it affect the performance? say both pointing in opposite direction...

+1
+How will perform if antennas are sectors pointed in same direction?
Any field experience, someone?

not same direction... will be pointing them in 2 different directions.

worked for me for 2 links in different directions, one 5km and one 8km; speed ~70mbps aggregate but I used RocketM5

we are talking about Mikrotik platform....

I would guess it would work a lot like a SISO connection on either leg. But you are losing the multipath gains. In short it seems to me you win by conserving channels (2 p2p's on a single channel) but lose by not seeing any real mimo performance and throughput gains. So...a wash? Depends on your needs.

If you have gain to spare in both directions and could absorb the additional noise you'd suck up, what about a set of MIMO omni's?

-Nelson

in mikrotik's terms i can use each port for 1x1 MIMO?? what say??

I would guess it would work a lot like a SISO connection

What will be the settings for Band (A, only A, A/N), Channel width (20MHz, 20/40MHz HT Above ...), Wireless protocol (802.11, nstreme, nv2..) and HT (chain0 and chain1) to work like SISO?

I would guess it would work a lot like a SISO connection

What will be the settings for Band (A, only A, A/N), Channel width (20MHz, 20/40MHz HT Above ...), Wireless protocol (802.11, nstreme, nv2..) and HT (chain0 and chain1) to work like SISO?

You ask a lot of questions in a field that is only recently been explored and not a lot of experiances are around. Why not test some yourself?

[/quote]You ask a lot of questions in a field that is only recently been explored and not a lot of experiances are around. Why not test some yourself?[/quote]

Thank you for your "kind" answer. I see that you were also talking the same here
http://forum.mikrotik.com/viewtopic.php?f=3&t=55198
Anyway, you're right, there is nothing better to learn how things work yourself, but I think the purpose of this forum is to share experience by asking questions to somebody who has already tried some setup.

Best regards,

if you want both headers to work if 802.11a org is set, then you have to enable both HT chains for tx and rx. Then you can use so called antenna diversity and have interesting configurations, like omni on one header and grid on other to connect clients further away.

Ive done this with needing solid 180 degree coverage, used 2 120 degree antennas, one on H polarity and one on V polarity, but angled to point 180 degrees coverage, good 2x2 coverage in the beam, outer edges get good 1x1 coverage.

Ive done this with needing solid 180 degree coverage, used 2 120 degree antennas, one on H polarity and one on V polarity, but angled to point 180 degrees coverage, good 2x2 coverage in the beam, outer edges get good 1x1 coverage.

So, if I understand well, you have a section where both 120 sectors are overlapping and because they are both differently polarized you can use 802.11n dual chain clients here?
And then outside this section where both sectors are only are covering with their respective signal you have single chain 802.11n clients working? Either H-pol or V-Pol?

Or do you use dual pol clients everywhere where only the mid section clients get signals from both sectors in relative same strength and outside this beam section the client either have the H-pol chain coming in much stronger than the V-pol end and vice versa?

What is the assigned data throughput these clients are supposed to get?
What are the real connection rates of both type of clients?
Tell us what you exactly do. This is interesting stuff...

Ive done this with needing solid 180 degree coverage, used 2 120 degree antennas, one on H polarity and one on V polarity, but angled to point 180 degrees coverage, good 2x2 coverage in the beam, outer edges get good 1x1 coverage.

So, if I understand well, you have a section where both 120 sectors are overlapping and because they are both differently polarized you can use 802.11n dual chain clients here?
And then outside this section where both sectors are only are covering with their respective signal you have single chain 802.11n clients working? Either H-pol or V-Pol?


What is the assigned data throughput these clients are supposed to get?
What are the real connection rates of both type of clients?
Tell us what you exactly do. This is interesting stuff...

Yeap exactly this. Most of the coverage is 2x2, but around the 180 degree mark I can't see both chain clear (either horiz or vertical depending on which end of the 180 degrees Im at) So? Single chain clients either vert or horizontal. Works really well I have it deployed in 2 areas this way. Customers are either 1.5 or 3 megabits, Not very fast, but future ready. and can get excellent 360 degree mostly 2x2 coverage with 2 radios and 4 sector antennas.

if you want both headers to work if 802.11a org is set, then you have to enable both HT chains for tx and rx. Then you can use so called antenna diversity and have interesting configurations, like omni on one header and grid on other to connect clients further away.

Is this kind of diversity expected to work correctly also in 802.11n with both chains enabled and a SINGLE spatial stream?

Consider this simplified hypotetical setup:

1) n.1 AP in a central location with RB52Hn working in 5GHz-band/802.11n with both chains enabled in both TX/RX, two directional antennas (one for each card's chain) with same (or different, it doesn't matter to what I mean) polarization, pointed to different and non overlapping zones.

2) n.2 STATIONS, one for each of the coverage zones above, each with a SINGLE directional antenna connected to chain0 only (TX+RX) and chain1 left unused/disabled (here you may assume a RB711-5Hn as station). Each station has its own antenna pointed to the corresponding AP's antenna (with correct polarity).

In the above example, each chain of the AP's R52Hn will receive ONLY (well... almost) the signal from a single station and each station will receive ONLY (...) the signal from a single AP's R52Hn chain: is this expected to work correctly?

In other words, with both chains enabled, is the R52Hn able to handle a single spatial stream (MCS0-7) on both chains? Explicitly disabling MCS8-15 on the AP's R52Hn will make any difference in this case? Using 20 or 40 MHz channels would change things?

Let me repeat again: I'm talking about 802.11N, NOT 802.11A.

Thank you for your patience in reading.
rock