So I'm considering using Nv2 for my PtP and PtMP links; essentially at every wireless bridge in my WISP infrastructure. Does anyone have any advise regarding this?

How well does Nv2 play with other WISP infrastructure which uses 802.11x protocols?

One key advantage I was sold on is the time slot based transmission instead of the traditional CSMA approach which in theory should allow for more focus on the data transmission/latency.

Also, I read somewhere that Nv2 enabled APs would not be as 'distracted' than other 802.11x APs because it's a completely different language... or should I say protocol. When I think on it though, since they all use the same frequency range, talking another language or not, isn't noise... well noise regardless of the language?

The analogy I have in mind is being in a busy fish market where there are many other people talking and perhaps even shouting to negotiate transactions.. Does it really matter if you speak another language? You're still shouting over me!

I await your expert advise

I have been using NV2 since its introduction, now it's very reliable, at first glance latency looks to have increased with NV2 (this has been reducing with development) under light traffic but here is one it's real advantages, unlike 802.11 does not increase rapidly as more and more traffic is pushed through, Hidden nodes are a thing of the past with NV2 and of course much higher throughput, as regards co-location interference it's much better at reducing it's effects but not eliminating the interference which will impact on performance, always keep this in mind,

Good to know

I must say, NV2 does seem to give better performance. However, it also seems like it will create a problem for others (i.e. competitors) that share the same space. This "RF Splatter" and "always communicating" even when there is little to communicate might just piss someone off.

In theory it seems annoying for others sharing the same space, how annoying is it really in practice?

Good to know

I must say, NV2 does seem to give better performance. However, it also seems like it will create a problem for others (i.e. competitors) that share the same space. This "RF Splatter" and "always communicating" even when there is little to communicate might just piss someone off.

In theory it seems annoying for others sharing the same space, how annoying is it really in practice?

Its always good to remember the frequencies wireless AP use (exclude licensed PTP + PTMP ) are in licence exempt frequency bands
and others are entitled to use these frequencies for whatever, so we always have to tolerate (manage) interference generated by other devices using these frequencies,

Its always good to remember the frequencies wireless AP use (exclude licensed PTP + PTMP ) are in licence exempt frequency bands
and others are entitled to use these frequencies for whatever, so we always have to tolerate (manage) interference generated by other devices using these frequencies,

I will be sure to quote this

in some scenarios nstreme is better than nv2, so you should test this too

in some scenarios nstreme is better than nv2, so you should test this too

So I've seen with my research on this forum.

Channel/TDMA Period/Power all to be fine tuned

in some scenarios nstreme is better than nv2, so you should test this too

I used nstreme and changed over to NV2 but would like a example where nstreme is better (apart from latency)

in some scenarios nstreme is better than nv2, so you should test this too

I used nstreme and changed over to NV2 but would like a example where nstreme is better (apart from latency)

Nstreme works much better regarding tcp speed esp. with multiple hops. But it suffers from interference much more than nv2.

So... apparently Mikrotik isn't the only manufacturer using TDMA:

http://dl.ubnt.com/datasheets/airmax/UB ... X_TDMA.pdf

So how does Nv2 differ from what UBNT is doing?

So... apparently Mikrotik isn't the only manufacturer using TDMA:

http://dl.ubnt.com/datasheets/airmax/UB ... X_TDMA.pdf

So how does Nv2 differ from what UBNT is doing?

All PTMP Vendors implement a polling protocol on top of wlan. The first I've seen was WORP from Proxim.
Reason is that standard wlan is not designed for longer distance ptmp.

No vendor describes the exact implementation of his protocol so you've to test if it does the job for you.

Hi,

We have been using Nv2 in our Multipoint infrastructure almost 18 months, the first experiences were ugly, there were some problems with prior versions, by now is very stable, now we want to get more performance, not only in troughput but also in latency, noise avoidance and CCQ, there are best practices tips for the Nv2 configuration or coverage and RF design?

We are using RB433+R52Hn and Ubiquiti 20dBi 90° Sector.

Thanks.
Marcelo.

@marchelo78

Which antenna exactly? I'm researching PtMP (sector antennas) .

Also, are you using more than one antenna with that single routerboard? I believe you have the ability to use three radios with that board right?

in some scenarios nstreme is better than nv2, so you should test this too

I used nstreme and changed over to NV2 but would like a example where nstreme is better (apart from latency)

Nstreme works much better regarding tcp speed esp. with multiple hops. But it suffers from interference much more than nv2.

I totally agree and Nstreme suffers badly from co-location interference and causes OSPF routing issues, using NV2 is the better option
so far as Nstreme development has ceased for some time.

I have over 20 long range point to points using Mikrotik, and I'll say I've had very mixed results.

Using the Mtik bandwidth testing, NV2 mostly shows more throughput than regular Nstreme. However, if your using any routerboards with GB interfaces, then going back to 100M AP or any device, you may have major slowness and need to use Nstreme.

The problem has never been fixed and I believe because of the complicated-ness of it, its simply been on the backburner.

GB > NV2 PtP > GB WAN > 100M = up to 50% throughput loss!

I guess all this really says then is that Nv2 theoretically should be better. But really, in practice, it's best to test every protocol for every application; it might not be the best in your situation.

I have over 20 long range point to points using Mikrotik, and I'll say I've had very mixed results.

Using the Mtik bandwidth testing, NV2 mostly shows more throughput than regular Nstreme. However, if your using any routerboards with GB interfaces, then going back to 100M AP or any device, you may have major slowness and need to use Nstreme.

The problem has never been fixed and I believe because of the complicated-ness of it, its simply been on the backburner.

GB > NV2 PtP > GB WAN > 100M = up to 50% throughput loss!

I have never experienced any major slowness using NV2 and certainly not 50% throughput loss but maybe other issues could cause this

GB > NV2 PtP > GB WAN > 100M = up to 50% throughput loss!

Don't want to teach you to suck eggs or anything, but have you tried hard-setting the speed/duplex on the gbit:100M link?

to answer your questions, I had some curious interference problems on 2 links where switching from nv2 to nstreme fixed a problem when the link was reconnecting every 5 or 10 seconds.
And of course, generally the latency is much better with nstreme than nv2. An increase from 1ms to 3-4ms is visible. I also have two ospf dual links (one link only tx, the other one only rx - some kind of "emulated" ol` good nstreme 2 using n) that work better using nstreme than nv2.
My last version on the links is 6.19, didn't test anything new since that because it's pretty stable like this.