View Full Version : Results on short distance indoors
opampca
08-14-2009, 12:54 PM
As everyone is doing long distance tests, I am doind one indoors.
I use m2, with going thru about 5 thick 100 years old wall in a city.
Noise level is surprisingly low.
AP has a 19dbi VAGI antenna and client has 16dbi panel.
Gain is adjusted for a max of 36dbm EIRP.
Airmax is off since it drives the throughput down to 3mbps... (???)
Waiting for the press release.... :-)
With Bullet M2 to Bullet 2 (G), I achieve 20mbps using jperf.
With M2 to M2 (N), 20mhz, I achieve 27mbps
With M2 to M2 (N), 40mhz, I achieve 32mbps
With 10 parallel streams, It sum at about 70mbps....
What is weird, is that all streams are not equal and tend to form 2 groups with one having almost double BW than the other.
More tests coming..
opampca
08-14-2009, 01:13 PM
When I do M2 to M2 20mhz, with a Bullet2 associated to the AP, I still have same results, seems that the AP do not go down in G mode for the M2-M2 link...
I get 27mbps, wether or not the B2 is associated.
Of course the AP has to go down to 20mhz for the B2 to be able to associate.
UBNT-Mike.Ford
08-14-2009, 01:19 PM
When I do M2 to M2 20mhz, with a Bullet2 associated to the AP, I still have same results, seems that the AP do not go down in G mode for the M2-M2 link...
I get 27mbps, wether or not the B2 is associated.
Of course the AP has to go down to 20mhz for the B2 to be able to associate.
How about when you stream multiple data streams (say 4 for 20Mhz mode, B2M to B2M?)
Thanks,
opampca
08-14-2009, 01:35 PM
4 streams sum to 32mbps, with BM2-BM2 20mhz with B2 associated
10 streams sum to 34 mbps
50 streams sum to 35 mbps also....seems to be the limit
Would that be a good way to stress test AP ?
UBNT-Mike.Ford
08-14-2009, 01:50 PM
Are you using AutoACK?
Thanks,
Mike
opampca
08-14-2009, 01:53 PM
Are you using AutoACK?
Thanks,
Mike
No, It is set to the lowest, 22
Remember I am at 20 meters apart
UBNT-Mike.Ford
08-14-2009, 01:58 PM
Hello,
This is fine please test again with AutoACK timing, it should improve.
Thanks,
Mike
opampca
08-14-2009, 02:10 PM
I'm getting the same results...
But I am quite impressed with the results...
I'm indoors with lots of obstruction in the way...
I am very anxious to test that thru the trees, we do a lot of rural installations this summer and so far, we had to use 900mhz quite often...if we can avoid that , to get the same results....
opampca
08-14-2009, 03:21 PM
Now, with 40mhz I'm getting 40mbps M2 to M2
10 streams, getting 76mbps total
Maybe the time of day.. :-)
UBNT-Mike.Ford
08-16-2009, 11:18 AM
Now, with 40mhz I'm getting 40mbps M2 to M2
10 streams, getting 76mbps total
Maybe the time of day.. :-)
Hey,
Can you test with 4 streams EACH way...so running bi-directional traffic at the same time?
Thanks,
MIke
opampca
08-16-2009, 07:22 PM
Sorry, but one of my 2 unit is on a roof directed to the mountain (wild) that sits in the middle of the city. I was there today, but didn't get good results. Only 5Mbps on a 3km link...
After these tests, I might redo the indoor setup, I only have 2 units.