dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
1506
share rss forum feed

h21j3

join:2013-11-07
USA

Bridging internet over 10 miles

Hi,
This is my first post to a forum ever mostly because I am very interested in this topic. I would like to bridge an internet connection over 10 miles from my house to my parents house since Timewarner doesn't cover his area. I am getting the 100mbps plan and would like them to at least be getting 50mbps. I have the general idea on how to do this but I would like some advice. I am willing to spend up to $1000 to have this set up and functional. I was thinking about using Ubiquitis Airgrid M 16dbi radios. So this is how I see it in my mind.
1. A high quality router and poe adapter at my house and their house.
2. Airgrid Antennas mounted 11m high at both houses.
3. My antenna is set up as a transmitter and plugged into router lan port.
4. Their antenna is set up as receiver and plugged into router wan port.
5. Im using airlink to calculate the link. This is what im getting with both radios 11m high.
-75dbm signal strenth - This is good?
Noise floor : -96 dbm - I'm clueless to what this is.
Transmit CCQ: 80% - I also don't know what this is.
TX/RX Rate: 120mbps/120mbps
Airmax enabled - A quality feature of the radio?
Airmax quality: 63%
Airmax capacity: 40%

If I had to use more then 1 antenna would it go like this?

1 antenna at my house pointing to first hop location.
2 antennas at first hop location, 1 receiving which is directly connected to second antenna which is transmitting to 2nd hop or parents house.
If there isn't a second hop then 1 antenna at parents house receiving from first hop transmitter antenna.

Sorry if I sound ignorant in all of this I know I am a noob at all of this but I do have some basic knowledge of networking. Am I using the right antennas and is my idea good enough to pull off?



Anav
Sarcastic Llama? Naw, Just Acerbic
Premium
join:2001-07-16
Dartmouth, NS
kudos:5

Depends of straight line Line of Site connectivity and a certain width around that for optimal connectivity (especially at the halfway point). google maps help.



Inssomniak
The Glitch
Premium
join:2005-04-06
Cayuga, ON
kudos:2
reply to h21j3

Best to post coordinates of the sites.

10 miles with only 11 meters of height sounds like a problem unless both of you live on hilltops.
--
OptionsDSL Wireless Internet
»www.optionsdsl.ca


h21j3

join:2013-11-07
USA

1 edit

Their house - 40° 4.670', -83° 27.021'
My House - 40° 3.452', -83° 32.017'

Those are not my silos but I might have permission to mount the radios there. Are nano bridge m by ubiquiti better than the airgrid series?



Mellow
Premium
join:2001-11-16
Salisbury, MD
reply to h21j3

What does airlink show with the Rocket Dish?


h21j3

join:2013-11-07
USA

Rocket dish? With the Rocket M antenna its showing signal strength of -65db and transmit ccq of 100%. Airlink quality and capacity at 100%. Similar results with the Nano Bridge M



clarknova

join:2010-02-23
Grande Prairie, AB
kudos:7
Reviews:
·TekSavvy DSL

Aim for ccq and AMQ of 100%. You want your signal strength to be -70 or better for reliability. -60 to -50 is ideal.

Note that your actual throughput capacity is roughly half of the link speed. So the best link speed you will see with a pair of Airgrid is MCS7 or 150 Mbps. Actual throughput will more likely be 60-70 Mbps at best. If you want more throughput than that then get a pair of 2x2 MIMO radios such as the Rocket.

23dBi antennas on both ends is probably the lowest you want to go, but for a 10-mile shot I would play on the safe side and go with RD30 on both ends if your mount can take it. This will give you your full throughput and good reliability, all in under budget.

Note too that you're likely to run into less interference on the 5GHz band compared to 2.4.
--
db


h21j3

join:2013-11-07
USA

Hey, thanks for the info.
I shouldn't have said 10 miles, it is actually less than 5 miles in a straight path. So I probably won't need the rocket dish.

Can anyone confirm that I have the hop idea correct?



clarknova

join:2010-02-23
Grande Prairie, AB
kudos:7
Reviews:
·TekSavvy DSL

Yes, with two pairs of radios you can do two hops. The radios in the middle will just patch together by ethernet.

In your case however, with 5 miles line of sight you should have no trouble doing a single hop with a pair of radios. Have a look at this comparative marketing pdf:

»ubnt.ca/img/pdf/airmax_comparison.pdf

Generally speaking, any 1x1 radio on there, when paired with a sufficiently specced antenna, will pass about 60 Mbps over good line of sight and reasonably low RF interference. The 2x2 radios should do about double that.

My estimate is that for a 5-mile shot in the 5GHz band you're going to want about 25dBi antennas at both ends for full speed. You can corroborate that with the airlink site.

The Nanobridge M5 might be sufficient for full speed. The Powerbridge M5 with its extra output power is a safe bet. A Rocket with a 30dBi Rocket Dish is probably overkill but would also work. If you're happy with 50 Mbps then the Airgrid M5 is a viable choice. Again, the 23dBi might be sufficient, but the 27dBi is the safer choice. Any of the above mentioned radios in a pair with typical mounting hardware and cabling will fall under your $1000 budget. Some of them well under.
--
db



Anav
Sarcastic Llama? Naw, Just Acerbic
Premium
join:2001-07-16
Dartmouth, NS
kudos:5
reply to h21j3

Im just glad Clarkie is here. Great advice as always.



WHT

join:2010-03-26
Rosston, TX
kudos:5
reply to h21j3

Click for full size
As it would be your luck, you are shooting right through a grove of hmm...10 meter trees that are on a 10 meter hill...that happens to be close to the widest Fresnel width.

Even if you can use the 65 foot silo across the street from you, you still won't have clearance.


WHT

join:2010-03-26
Rosston, TX
kudos:5

1 edit

1 recommendation

reply to h21j3

UBNT Radios
/edited to mask out dual chain for 5.8 airGrids


clarknova

join:2010-02-23
Grande Prairie, AB
kudos:7

Nice table.
--
db



WHT

join:2010-03-26
Rosston, TX
kudos:5
reply to h21j3


h21j3

join:2013-11-07
USA
reply to clarknova

Thanks for the informative reply. It cleared up most of my questions. I'll probably end up getting a couple of the Powerbridge m5s as you recommended and 1000 feet of Ubiquitis Tough Cable.


h21j3

join:2013-11-07
USA
reply to WHT

You're saying even with the Rocketdish it won't work? I have heard of Nanobridges having a decent signal through a small woods, so couldn't the powerbridges work?


ArizonaSteve

join:2004-01-31
Apache Junction, AZ
Reviews:
·voip.ms
·CenturyLink
·T-Mobile US
reply to h21j3

The Know How show on twit.tv had an episode #38 showing just how to do that. The distance is shorter but the principle is the same. However you can get distances up to 30 miles if you have enough power and big antennas so you shouldn't have to make any hops unless there is a hill in the way. twit.tv/show/know-how/38



WHT

join:2010-03-26
Rosston, TX
kudos:5
reply to h21j3

Good luck with your project (code for "you're ignoring the advice from professionals who do this daily"). I'm done here.


LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1

Programs such as Radio Mobile are good for profiling the terrain but their landcover is not always accurate. While your link might be able to tolerate a bit of fresnel incursion by foliage, it is not something that can be quantified. I can tell you that over that distance, every dB counts.

Bring up the link in Google Earth and see what is along the path. If you have a GPS put the two endpoints in and then go for a drive and a walk to see if the sat image still reflects reality. Lastly, climb up on the silo at dusk and have someone flash the far end.
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey



WHT

join:2010-03-26
Rosston, TX
kudos:5

1 edit

Radio Mobile land cover is accurate. Well hashed over in the UBNT forum.
/edited Though you should calibrate the default values to fit your environment.

Foliage land clutter reduces the impact of Fresnel Zone problems.


LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1

said by WHT:

Radio Mobile land cover is accurate. Well hashed over in the UBNT forum.

You can stick a fork in it and call it done if you want but that doesn't mean I agree. There is no friggin' way you can call out a single hedgerow of trees from years old sat imagery. It could be better or it could be much worse.

10 miles is a tough shot with such a narrow margin on the fresnel zone. Nobody will give a money back guarantee. Buy the kit and try it. If it doesn't work, there's always ebay.
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey


Anav
Sarcastic Llama? Naw, Just Acerbic
Premium
join:2001-07-16
Dartmouth, NS
kudos:5
reply to h21j3

your plan of hop in-between if you can find clear pathways may also be feasible (can you find a third useable silo??


LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1
reply to LLigetfa

said by LLigetfa:

said by WHT:

Radio Mobile land cover is accurate. Well hashed over in the UBNT forum.

10 miles is a tough shot with such a narrow margin on the fresnel zone.

Not sure why the topic title says 10 miles when your RM plot says 5 miles. That certainly reduces FSL improving the odds.

5 miles also makes it easier to visually verify the path.
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey

h21j3

join:2013-11-07
USA
reply to WHT

I wasn't ignoring your advice, I am open to all suggestions. It would be Ok with me if this project doesn't work out too.


h21j3

join:2013-11-07
USA
reply to Anav

There are a lot of silos around me so I could probably find something that would work. In that case I wouldn't get 2 power bridges but probably 4 Airgrids or Nanobridges.



WHT

join:2010-03-26
Rosston, TX
kudos:5
reply to LLigetfa

I guess the OP over estimated the distance. My 7 KM RM plot is based on the locations he posted. Imagery can resolve a single large canopy tree. Examples were given in the WISP forum a few years ago comparing imagery to what was really on the ground.


LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1

said by WHT:

Imagery can resolve a single large canopy tree. Examples were given in the WISP forum a few years ago comparing imagery to what was really on the ground.

My experience with RM and landcover data in my area yield very different results. YMMV.
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey


Mellow
Premium
join:2001-11-16
Salisbury, MD
reply to LLigetfa

said by LLigetfa:

Lastly, climb up on the silo at dusk and have someone flash the far end.

Would that work at this range? More curious than anything, are you talking about using a high powered flash from a camera or a high powered laser?

LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1

A camera or strobe light. If there is high ambient light at the other end, flash from the silo.