There are some old topics on link modeling, but I thought I'd start a new one rather than pull up the old ones. I'm attempting to model a PtMP infrastructure using RadioMobile Online and I'm having some problems wrapping my head around some basic pricipals. Help a newbie out, woudja?

The Nanostation M2 has specs for MCS index data rates as well as 11b/g data rates. I've read up on what the acronyms mean, and I'm just starting to get a grip on the subject.

I get the impression that in a MIMO radio, the radio (and it's driver) make some decisions on what modulation scheme and other parameters to set, and that's reflected in the MCS index.

So here's the basic question... Does BBHN preserve the radio's MIMO operation, making the MCS specs relevant, or does BBHN lock the NSM2 into a 802.11b/g mode which locks the radio out of using MIMO parameters?

I've empirically measured a link of 13 km (9 miles, more or less) using a pair of NSM2's with clear line of sight over an urban environment. BBHN V1.0 s/w.  Data rate was not great, but adequate; around 3-5 Mb/s. Some drop-outs no doubt caused by noise; the GUI reported a S/N ratio of 0-5 db, which I thought showed great performance in a harsh environment. Rx level indicated by the GUI was bouncing around -80 to about -85.

I've found a post that says to use a rx threshold of 4.74 uV in radiomobile. I believe the post was in reference to Linksys radios. Is this a valid value for the NSM2? That's about -75dbm according to my calculator.

Thanks for any edification offered.

The chipset detetmibe a what mode to operate under.  

It should I belive operate up to MCS7.  The chipset /driver makes the decision about what mode to go into.  In your case the noise floor is the limiting factor. the nodes have to push higher power (slower speed) to get over the noise. 

As speed picks up the tx power decreases (linearity of signal amplifier) so you can't expect full speed in your case because it would drop the signal into the noise floor.

for modeling levels:

personaly I model anywhere from -96 to -85 dbm  and extrapolate based on what I know. Modeling is a bit of an art as much as if is a science you have to know when to doubt the data and when to trust the data.

a -74 would be a generally safe number in many cases as that gives you plenty of room (~22db) for noise floor and for signal strength for a good strong signal but would likely be overkill for a true "I can connect and use it" model. 

Ive seen some nodes flirting in the -86dbm range (noise floors in the -92ish). Are they my best links? No , do they connect and get the job done, yes,  would I like better, of course.

Conversely I got another couple nodes in the -74dm range with a -82dbm noise floor (yes it hurts ) and they have some fair amount of drop (30-50%)   If I put some small dish antennas it will get better I just haven't done it yet. 

So the -70's won't hurt to model in but know it's possible to do in the weaker zones.

I tend to model using a low gain antenna (10dbi) due to my conservative nature and know I can make the link stronger if needed that is one reason I model closer to the -96 mark but I also know not to assume any link is guranteed to work..

For those using RadioMobile for modeling, here's a link to the Ubiquiti antenna pattern data files.  The ".ant" files can be uploaded directly to RadioMobile Online.

http://community.ubnt.com/t5/airMAX-Frequently-Asked/airMAX-Do-you-have-antenna-patterns-available/ta-p/902862

Remember, when you use a dual polarization antenna (such as UBNT NanoStation, one of their various sector antennas, RocketDiish, etc.) you're actually using two antennas... each having their own V and H patterns. They'll need to be modeled separately.  Their file naming convention describes the model and polarization the data file represents.

Andre, K6AH