Reply
Established Member
Posts: 1,536
Registered: ‎08-06-2010
Kudos: 308
Solutions: 3

LTU and Interference

How does LTU handle heavy interference in an urban environment?  I have a site in a big city where Airmax AC tends to fall flat.

Emerging Member
Posts: 102
Registered: ‎09-15-2016
Kudos: 25
Solutions: 1

Re: LTU and Interference

It still gets hit pretty hard even with the filtering stuff. You will definitely want to buy an RF armor set (box and dish shield) to minimize interference. 

SuperUser
Posts: 5,892
Registered: ‎08-26-2009
Kudos: 1788
Solutions: 58

Re: LTU and Interference

The ISO-BEAM sheild is probably cheaper.

Senior Member
Posts: 2,555
Registered: ‎03-23-2008
Kudos: 508
Solutions: 18

Re: LTU and Interference

[ Edited ]

@twinkletoes wrote:

The ISO-BEAM sheild is probably cheaper.


Of course it's cheaper, because it doesn't include a well designed and built shield box that blocks the RF getting into AF-5XHD through the plastic half of the radio case.  Has made a 1 to 3dB reduction in interference from back-to-back, co-located radios between having the RFArmor box lid on or off.

 

On R5AC-Prism with "full-metal-jacket" (front and back of case), the improvement is only 1 to 2dB.

 

If we are talking connectorized dish antennas, on RD-5G30 & 5G34, ISO-Beam is not compatible, so only the RFArmor option works.

 

The ISO-Beam 2' (620mm) dish shield is the only choice for the "LW-style" antennas. But in those cases we still add the $50 RFArmor Box.

 

Because given the convenience of not having to use "insurance" rubber tape on the pigtails, and the cleanliness of the connections inside the RFArmor box to allow a quick failing radio swap, I have never regretted spending that extra few dollars. Biggrin

 

Best Regards ... Joe

If the communication industry had been built on the backs of yes-men,
we would be submitting our forum posts at the telegraph office in town.
Senior Member
Posts: 3,758
Registered: ‎01-13-2009
Kudos: 705
Solutions: 9

Re: LTU and Interference

From what I hear, LTU actually falls flat on its face with any interference, similar to LTE-U on 5Ghz. Why didn't UBNT implement ANY type of noise filtering with LTU???? Nothing?? Unbelievable... I expected at least PRISM type filtering
Veteran Member
Posts: 5,669
Registered: ‎07-03-2008
Kudos: 1775
Solutions: 137

Re: LTU and Interference

Even at this (early) stage of firmware development, LTU has far better interference rejection than the AF-5X did.  We've seen ~60% improvements in capacity upgrading links in congested areas.

 

LTU filtering is in a completely different league than PRISM, and again -- there is stuff in the chipset that isn't even lit up yet.

SuperUser
Posts: 16,315
Registered: ‎02-03-2013
Kudos: 8928
Solutions: 595
Contributions: 2

Re: LTU and Interference


@wispwest wrote:
From what I hear, LTU actually falls flat on its face with any interference, similar to LTE-U on 5Ghz. Why didn't UBNT implement ANY type of noise filtering with LTU???? Nothing?? Unbelievable... I expected at least PRISM type filtering

The LTU radios (and also the older AF5X), have much, much betteradjacent channel filtering than the AirMax AC Prism radios.  Every LTU PTMP CPE radio has adjacent channel filtering better than AirMax AC Prism.  

ubiquiti certified trainer :: ubwa | uewa
Ubiquiti Employee
Posts: 6,512
Registered: ‎05-13-2009
Kudos: 2044
Solutions: 198

Re: LTU and Interference


@wispwest wrote:
From what I hear, LTU actually falls flat on its face with any interference, similar to LTE-U on 5Ghz. Why didn't UBNT implement ANY type of noise filtering with LTU???? Nothing?? Unbelievable... I expected at least PRISM type filtering

Unbelievable. I even don't know what to say more here... Have you only heard or have you also tried to replace any PRISM link?

Regular Member
Posts: 636
Registered: ‎10-11-2013
Kudos: 184
Solutions: 3

Re: LTU and Interference


@UBNT-Edmundas wrote:

@wispwest wrote:
From what I hear, LTU actually falls flat on its face with any interference, similar to LTE-U on 5Ghz. Why didn't UBNT implement ANY type of noise filtering with LTU???? Nothing?? Unbelievable... I expected at least PRISM type filtering

Unbelievable. I even don't know what to say more here... Have you only heard or have you also tried to replace any PRISM link?


I second this. Critics should be paired with facts.

 

Problem is that there are no vendor data out regarding interference behaviour. ClaudeSS did a great job with a test setup (which is outdated now ;-)) testing real behaviour of different vendor equipment. This matched my experience. My AF5XHD links run fine. There is room for improvement handling the presence of interference. We have only DFS channels so we would love to see some efforts avoiding this 60s delays wherever possible. My dream is a smootheless migration between channel sizes and channels while the traffic flows.

Senior Member
Posts: 3,758
Registered: ‎01-13-2009
Kudos: 705
Solutions: 9

Re: LTU and Interference


@UBNT-Edmundas wrote:

@wispwest wrote:
From what I hear, LTU actually falls flat on its face with any interference, similar to LTE-U on 5Ghz. Why didn't UBNT implement ANY type of noise filtering with LTU???? Nothing?? Unbelievable... I expected at least PRISM type filtering

Unbelievable. I even don't know what to say more here... Have you only heard or have you also tried to replace any PRISM link?


I have several, probably 6 links in place now that upgraded from legacy AF5X. 

 

It was an improvement in capacity, but noticed when modifying channel size and/or when changing to 50/50 split, both ends of the spectral graph would have to be perfect green. If one side even had a slight color change in it, the capacity would drop big time. Also larger channel sizes over 50mhz yielded less throughput than smaller, again simply because the spectrum isn't 100% perfect. 

 

Then on FB group posts one of your alpha testers mentioned it doesn't have any noise filtering, mainly because it's not 802.11 product. 

 

I know LTU is good for blocking out noise that's outside of it's own channel, but is there really nothing that can be done to help with 802.11 noise from within a desired channel?

Regular Member
Posts: 636
Registered: ‎10-11-2013
Kudos: 184
Solutions: 3

Re: LTU and Interference


@wispwest wrote:

@UBNT-Edmundas wrote:

@wispwest wrote:
From what I hear, LTU actually falls flat on its face with any interference, similar to LTE-U on 5Ghz. Why didn't UBNT implement ANY type of noise filtering with LTU???? Nothing?? Unbelievable... I expected at least PRISM type filtering

Unbelievable. I even don't know what to say more here... Have you only heard or have you also tried to replace any PRISM link?


I have several, probably 6 links in place now that upgraded from legacy AF5X. 

 

It was an improvement in capacity, but noticed when modifying channel size and/or when changing to 50/50 split, both ends of the spectral graph would have to be perfect green. If one side even had a slight color change in it, the capacity would drop big time. Also larger channel sizes over 50mhz yielded less throughput than smaller, again simply because the spectrum isn't 100% perfect. 

 

Then on FB group posts one of your alpha testers mentioned it doesn't have any noise filtering, mainly because it's not 802.11 product. 

 

I know LTU is good for blocking out noise that's outside of it's own channel, but is there really nothing that can be done to help with 802.11 noise from within a desired channel?


You cant filter on channel noise. For sure equipment differ on how much bandwidth to achieve with a given SNR. You may increase the SNR with the slanted antenne a bit. What works for Wimax/LTE/5G is the adaptive usage of subchannels. So the interference is reduced by just not using the parts where the interference shows up.

 

Dont know what is in LTU chips now, but it was announced that there will be distinct channels with one radio possible. So it might be possible to work around interference with future SW/HW Releases.

  

SuperUser
Posts: 5,892
Registered: ‎08-26-2009
Kudos: 1788
Solutions: 58

Re: LTU and Interference

Clearly, since 802.11 noise is not real noise, it's just 802.11 noise, I'm sure we can just pretend it's not there and LTU will work perfectly. Reading more, it says right in Shannon's Theorem from 1949 that "802.11 noise does not affect the maximum rate that information can be transmitted within a communications channel and is not actual noise at all." 

 

Just kidding, it actually defines the limit of a communications channel with noise, including noise from 802.11 devices. Although it's from 1949, these kind of things become fundamental defintions of our physical reality, until proven wrong. Noise is noise and the rules here apply to everyone, no matter what kind of hardware you have. Nobody escapes these fundamental rules.

 

Yes LTU has no PRISM filter - it uses more sophisticated methods available from the ground-up design to achieve the same level of adjacent channel interference rejection.

 

The PRISM filter can't help you with noise within your same channel. But the function it does provide has an equivalent in the LTU hardware.

 

 

SuperUser
Posts: 13,923
Registered: ‎12-08-2008
Kudos: 10693
Solutions: 658
Contributions: 1

Re: LTU and Interference

[ Edited ]

Right - IIRC Bill P told us in the original LTU announcement  that the LTU chip had 4 or more individual filters under the control of the CPU which are dynamically used to remove interference as needed.   So way more control and power than Prism.

 

Here's a story I posted about the real-world gains we saw on a 21 mile link to a very noisy site (over 20 APs on 5GHz from another ISP) and the direct improvement in upgrading from an AF5X to LTU/AF5XHD.   And there was a large increase on it when we moved from AC Prism to AF5X years ago too...

 

https://community.ubnt.com/t5/airFiber-Stories/Long-distance-very-noisy-AF5XHD-much-better-performan...

 

Prism is very good at what it does, but it uses up/down conversion to get to the baseband where the filter does it's thing, and that adds noise into the system overall.   So it's a balancing act - add noise from the up/down conversion vs remove noise with the filter (which also has loss itself).   It's not as simple as "let's add a filter" or everyone would do it...  ;-)

 

Jim

" How can anyone trust Scientists? If new evidence comes along, they change their minds! " Politician's joke (sort of...)
"Humans are allergic to change..They love to say, ‘We’ve always done it this way.’ I try to fight that. "Admiral Grace Hopper, USN, Computer Scientist
":It's not Rocket Science! - Oh wait, Actually it is... "NASA bumper sticker
":The biggest problem in tech I see right now is that most users don't want to do things that are hard. That doesn't bode well for the industry or the society.": (me. actually ;-)
Senior Member
Posts: 3,758
Registered: ‎01-13-2009
Kudos: 705
Solutions: 9

Re: LTU and Interference

First off, when I mentioned 802.11 noise I was just referring to 802.11 as that's most of what the interference is on 5ghz. I wasn't specifically stating LTU can magically ignore it, etc lol

 

I just see all these claims from Cambium and Mimosa, ets about noise filtering techniques, "noise immunity", etc. then UBNT has the PRISM filter, etc. 

 

Then I noticed that on every AF5X-HD I installed, the slightest little bit of noise would tank the capacity, big time. Then I found out there's no filters, then I saw this thread here, and it all came out as it did.

 

It just really makes me nervous thinking of using this in PtMP. Someone even mentioned that the Airmax AC will perform better than LTU in noisy environments. 

 

As someone else mentioned, if LTU could be similar to LTE and other robust products and utilize 100's or 1000's of sub-carriers, it can mitigate and "cut out" the slices of spectrum are noisy rather than drag down the entire channel

SuperUser
Posts: 5,892
Registered: ‎08-26-2009
Kudos: 1788
Solutions: 58

Re: LTU and Interference

Well see, there you are assuming that LTU has "no filter" again, but it has some advanced filters!

 

So, the ability to individually measure the interference at subcarriers and cancel out the worst ones is interesting, and perhaps someone can comment on LTU's ability to do this, now or in the future??

SuperUser
Posts: 5,892
Registered: ‎08-26-2009
Kudos: 1788
Solutions: 58

Re: LTU and Interference

I think it's worth saying, LTU is a very high modulation rate system. High modulation rates are inherently unstable and prone to interference. It's easy to get any high modulation rate system to tank. How it handles interference relates to filters, features like barring out bad subcarriers, or even modulation rate control, or any of the other modern techniques available, how these features are handled by the LTU team is going to determine how LTU stacks up against other high modulation rate systems...

Member
Posts: 287
Registered: ‎10-26-2009
Kudos: 57
Solutions: 2

Re: LTU and Interference

We have replaced AF5x links with HD's and seen very nice improvements in throughput. Every 8x link is easily 10x now and some that were 6x are also a stable 10x. Of course you need a fairly clean channel and good enough signal and SNR to reach 10x but I am seeing excellent performance from these radios even with very close adjacent channel interference.

 

Please STOP assuming a radio can handle heavy interference when you're on the same channel. Move things around to find a better channel as this still isn't magic. Overall, I find these radios handle inteference better than Prism, 5X, and definitely better than Mimosa does. Anyone who says this radio performs worse than a 5X or Prism on the same channel and channel width has something misconfigured in my opinion.

Highlighted
Senior Member
Posts: 3,758
Registered: ‎01-13-2009
Kudos: 705
Solutions: 9

Re: LTU and Interference

Agreed. Every 5X I ugraded, I had to really spend the time fine-tuning the center channel and size. 

 

One link had a slight color of interference on the upper side of the channel, and the other end of the link had some on the lower side of the channel. I basically had to "sandwich" the channel right in between and it works great, almost double what I had before. 

 

HOWEVER, I couldn't imagine trying to fine tune this on PtMP mode with 30-40 customers attached. What's going to happen then..

Member
Posts: 287
Registered: ‎10-26-2009
Kudos: 57
Solutions: 2

Re: LTU and Interference

You have to remember that the customer side should have less interference because it is 15-20 feet off the ground and not a PTP dish on a tall tower like your backhaul links. It's an entirely different story but how do you envision this being any different than current PtMP setups today? We have an AP on a 90* sector or something more narrow like 30* horns.

 

We'll still have to pick clean channels like we do today and the tower AP will still hear more noise than the CPE will. Having realtime spectrum analyzer will be helpful and the HD is the best analyzer I've seen so far and can cancel out in-band noise from it's own radios on the connection.

 

So this shouldn't be any harder finding a clean channel than what we already have to do today. I think you're really blowing this out of proportion. I am no fanboy and I agree PtMP is way behind schedule from what they've told us over the last 2 years BUT if they're making sure it's fully baked before launching, that is a great thing that Ubnt hasn't done well in the past. So I give kudos to them for waiting until it's ready, kind of like Cambium does historically. Although the Force 300 and ePMP 3k have also been delayed as long as LTU PtMP has due to stability issues too.

Senior Member
Posts: 3,758
Registered: ‎01-13-2009
Kudos: 705
Solutions: 9

Re: LTU and Interference

We have a tower with 16 horns (30 degree) and 500 clients. It's getting a bit rediculous. Others are similar but more like 6-10 horns. 

 

Most customers all have dual-band routers now that throw out 5ghz noise. It's everywhere now. What if customers dish shoots over neighborhoods loaded with 5ghz AP's?

 

Your right, no difference than now with Airmax AC, except to really see any throughput benefits with LTU you need really clean spectrum at both ends. Otherwise if noise drops it to AC or lower speed, what's the point? From my 5-6 HD PTP links it just felt like it was REALLY picky on making sure both sides were 100% clean and in the green. Otherwise capacity would tank down to regular AF5X or worse. 60mhz channels in somewhat clean spectrum would yield horrible throughput capacity, 80-100 wouldn't even register. 

Reply