What do I need to complete this network?

[ron_jeremy]

I am trying to finish someone else's work.

The job is to provide an Internet connection to an island property that is about 1 mile off-shore.

There are no devices on the transmitting side of this network; all devices are on the receiving (island) side.

I am unsure about whether I need 2 routers or 1 router + 1 access point, and whether I also need to use a bridge or 2 (wired or wireless?).

The antennae are already mounted and aligned.

This layout drawing is what I'm talking about:

http://i72.photobucket.com/albums/i181/thanasi67/05%20proposed%20wireless%20network_zpswldlngjd.jpg

 

[stevech]

forget the antennas. N/A.

get a pair of these. mount them at 30 ft or more.
http://www.newegg.com/Product/Produ...&cm_re=engenius_bridge-_-33-168-114-_-Product

Beware salt spray and corrosion. You'll need the proper weatherproofing tape, etc. And experience on how to avoid seaside short lived electronics. Else you'll be back at it after a storm.

 

[ron_jeremy]

Thanks for the input, but none of my questions were addressed...

About the antennae; they were used by the previous owner but left behind (the electronics are all gone). Both antennae are in impossible (for me) to reach high places, so I will (hopefully) not be touching them.

I simply need to know what hardware is required for all this to work.

 

[ron_jeremy]

Please note: the schematic I drew incorrectly lists the distance as 1 mile. I should be 3.5 miles. I will try and correct it.

 

[stevech]

I simply need to know what hardware is required for all this to work.

I suggest you use the hardware I showed in the post, above. A pair of 5.8GHz bridges with integral antennas.

But now, you say the path length is 3.5 miles. This changes things. You'll need a pair of outdoor bridges such as Cisco makes mounted adjacent to a 24dBi dish.
This needs a professional to engineer, furnish and install.
The above assumes use of unlicensed bands.

If your requirements are for more reliability and many users on the island, you may need a system in the price range of $10K and uses, say, 22GHz self-licensed. Depends on where in the world this is and the local regulations.

Hire a local subject matter expert or one here in the US to whom you can provide all the necessary details on requirements.

 

[ron_jeremy]

Thanks for the info. I just found out that the equipment used by the previous owner was Aironet BR2040-EE used in conjunction with the 2 existing antennae.

I can have it if I want it. I'm told it's a good 10 years old. What should I do?

 

[Cloud200]

I strongly suggest you start asking on a forum that is geared to WISPs.

Ie. Mikrotik and ubiquiti
http://forum.mikrotik.com
http://community.ubnt.com

 

[azazel1024]

You are likely going to have to tweak things to work with 802.11 long link issues.

I unfortunately can't view the images, but if you say that the antennas are mounted in impossible to reach locations, you'll likely need/want to make it reachable. That to me says several meters of coax, or more. High frequency in coax losses signal strength RAPIDLY. Figure about 2-3dB of loss per meter of length, which means if the cabling is 5m down a pole to where you could plug in a router/bridge, you are going to have 10-15dB of signal loss, which is significant, especially if you are trying to connect at a distance of 3.5 miles.

You are going to want an all in one bridge at the top of the pole, or a bridge plugged in to the antennas mounted at the top of the pole/location so that you only have a few feet of coax to run, then run an outdoor rated ethernet cable down the pole in to whatever router you want to use.

 

[stevech]

Thanks for the info. I just found out that the equipment used by the previous owner was Aironet BR2040-EE used in conjunction with the 2 existing antennae.

I can have it if I want it. I'm told it's a good 10 years old. What should I do?

Like
http://www.ebay.com/itm/Aironet-BR2...Multipoint-Wireless-Bridge-Used-/251515152754

This is a point to point bridge much like I spoke of above.

If I recall correctly, Aironet was a startup in early WiFi, early 1990's. It was acquired by Cisco (not Linksys). Cisco Aironet sold zillions to the professional market. I and many others used it for enterprise projects. Cisco today still sells successor products rooted in that Aironet line. It is, in my view, the best WiFi for professional use. They sell a lot of newer versions of bridges like shown above. Often, they are connected with a 1 ft. coax to a small dish for multi-mile bridge links, usually in 5.8GHz, unlicensed ISM band.

At 3.5 miles, this is ALL ABOUT THE TYPES OF ANTENNAS USED. And setup an ordinary spreadsheet link budget for a point to point link. I did this for a living along with other wireless projects, until mid-2014 when semi-retired. There are lots of people around who know how to do this.

I really think you need a consultant to help.

 

[sfx2000]

You are likely going to have to tweak things to work with 802.11 long link issues.

I unfortunately can't view the images, but if you say that the antennas are mounted in impossible to reach locations, you'll likely need/want to make it reachable. That to me says several meters of coax, or more. High frequency in coax losses signal strength RAPIDLY. Figure about 2-3dB of loss per meter of length, which means if the cabling is 5m down a pole to where you could plug in a router/bridge, you are going to have 10-15dB of signal loss, which is significant, especially if you are trying to connect at a distance of 3.5 miles.

You are going to want an all in one bridge at the top of the pole, or a bridge plugged in to the antennas mounted at the top of the pole/location so that you only have a few feet of coax to run, then run an outdoor rated ethernet cable down the pole in to whatever router you want to use.

At 3.5 miles you start running into timing issues with the 802.11 framing - there are a couple of companies that do a modified 11n radio where they have changed the MAC to a TDM format for bridging long links...

Ubiquity, I think, did this, might still - and there are others whose name escapes me at the moment.

sfx

 

[azazel1024]

DD-WRT and OpenWRT AFAIK allow you to change the frame retransmit timing. I'd imagine some of the high powered bridges out there allow you too as well.

Elsewise you are limited to a maximum range of 1 mile/1.6km based on the delay that is imparted to allow an acknowledgement of the packet(s). Longer range than that will cause problems/will not work.

Aironet stuff I am sure had the ability to change this as well.

You can also get reflections from the surface of the water and other issues at that kind of distance. I am sure it is perfectly workable, but it won't be super easy.

 

[ron_jeremy]

Thanks for all the help.

I guess I should make it clear that the current 10+ year old setup worked perfectly for years until very recently when the Aironet BR2040-EE was removed. Both the cantenna and receiving antenna are exactly where they've been for the last dozen years or so. The former owner said the network cost "thousands" to build at the time.

So, my question is:

• I now have the Aironet BR2040-EE in my possession, but am not sure whether I want to (re)install it or use a modern equipment. There is another one like it already in the house and it's connected to the roof mounted dish antenna (the receiving side). I have no issues replacing both Aironet pieces with equivalents from EnGenius, Ubiquity, Mikrotik, or whatever else.

 

[ron_jeremy]

One more thing: when I get a chance, I'll snap a photo of both antennae to give a better view of the equipment being used.

 

[azazel1024]

What was "working fine"?

In vague theory, replacing it with more modern stuff, so long as it really can be setup for a long link, should provide significant increases in performance, so long as the link budget was/is sufficient to do higher modulation rates.

That old 2040EE appears to be pre-802.11b. The PDF user manual claims it supports transfer rates of "up to 4Mbps" which is rather underwhelming, but at the same time, depending on HOW it was doing it, it also means that it was quite likely able to do that over rather extreme distances (the user manual also claims up to 25 miles).

So a newer bridge might not actually get you better performance, mostly depends on that link budget. You might need a more powerful antenna on the receiving end on shore, which might also be an advantage by reducing diffraction or reflection off the water.

Considering the relatively inexpensive cost (all things considered) I'd certainly look in to a couple of the less expensive wireless bridges out there that take a suitable external connector to connect up the bridge to the antennas and try it out. I'd just make sure you have a good return policy on the equipment in case it does not work out at all.

I would think you should at least be able to step it up to 11b rates or maybe even 11g rates at "only" 3.5 miles. Its a long distance, but with even a couple of good Yagi's on each end I'd think 20-30Mbps should be doable if not better. In bad weather though that might be a different story.

 

[ron_jeremy]

That old 2040EE appears to be pre-802.11b. The PDF user manual claims it supports transfer rates of "up to 4Mbps" which is rather underwhelming, but at the same time, depending on HOW it was doing it, it also means that it was quite likely able to do that over rather extreme distances (the user manual also claims up to 25 miles).

I am confused as to how a bridge can affect how far the signal goes. Isn't that all up to the antenna?

 

[stevech]

I am confused as to how a bridge can affect how far the signal goes. Isn't that all up to the antenna?

Bridge, AP, WiFi Router... "how far the signal goes" is naive.

• Which modulation mode (bit rate on WiFi, not IP layer)
• Transmitter's power for that modulation mode (it varies)
• Loss in antenna cables, if any
• Antenna gain for the boresight angle of interest
• Receiving site's antenna gain at same angle
• Receiving site's coax losses if any
• Receiving site's required signal strength for the modulation mode in use, for packet error rate of x% (typically 2% or less)
• Benefit of MIMO, if any
• Path length
• Frequency (e.g., 2.4GHz)
• Antenna height (affects Fresnel zone which affects path loss)
• Path loss for line of sight, without Fresnel zone attenuation, the path loss is a simple calculation from physics
• Additional path loss due to Fresnel zone impairments OR: obstructions in the path (trees, terrain, buildings, etc.) This is a large factor.
• Desired fade margin (excess signal strength)

Together, these numbers are a "Link Budget".
(We) RF engineers calculate this for every project to determine last number. We use our own spreadsheets or calculator tools from vendors.

Sorry, but the TRUTHFUL answer to How Far Does It Go comes from the above. Or build and trial and error. Doing the math, we know if it'll work before it's built.

 

[Samir]

Thanks for all the help.

I guess I should make it clear that the current 10+ year old setup worked perfectly for years until very recently when the Aironet BR2040-EE was removed. Both the cantenna and receiving antenna are exactly where they've been for the last dozen years or so. The former owner said the network cost "thousands" to build at the time.

So, my question is:

• I now have the Aironet BR2040-EE in my possession, but am not sure whether I want to (re)install it or use a modern equipment. There is another one like it already in the house and it's connected to the roof mounted dish antenna (the receiving side). I have no issues replacing both Aironet pieces with equivalents from EnGenius, Ubiquity, Mikrotik, or whatever else.

I would start by finding out why it was removed in the first place if it was working to deem if it should be re-installed. If there was no good reason to remove it, and there weren't any issues, I would install it again and then start researching an upgrade if they are looking for one.

 

[stevech]

Could it be that these were removed because they are at best 4Mbps because of the state of the art way back when these were designed?

Someone was hired/volunteered to replace this with higher speed gear, but that work was unfinished or being done by someone lacking the proper RF engineering skills for a 3.5mi. bridge link?

 

[azazel1024]

Could it be that these were removed because they are at best 4Mbps because of the state of the art way back when these were designed?

Someone was hired/volunteered to replace this with higher speed gear, but that work was unfinished or being done by someone lacking the proper RF engineering skills for a 3.5mi. bridge link?

I would guess the later.

I would also go with reinstalling it for now and researching/looking for new gear. In theory, if it is working with the antenna and wiring setup now, a newer bridge of roughly similar radio power and amplifier distortion SHOULD be able to manage a faster link rate. That is assuming that the old setup did not have marginal link budget.

However, I'd assume here that it was probably good. 3.5 miles is pretty far, but if it is line of sight with decent antennas on both ends, it isn't really that far. Its just long link in terms of how you must setup the bridge radios to ensure it can actually connect (because otherwise you run in to issues where the bridges are retransmitting packets because they didn't wait long enough to receive acknowledgement from the other bridge).

A decent router to router bridge with 100mw radios and 3dBi omni antennas can easily connect with 65Mbps modulation rate at a distance of 200ft line of sight. 3.5 miles is a LOT further away, but if you assume 4x the transmit power (400mw is a medium powered dedicated bridge), that gets you similar modulation rates at twice the distance (inverse square law), or 400ft.

Moving from 3dBi antennas to 16dBi antennas (medium gain yagi) on both ends is 26dBi higher gain between the two sides. That is roughly 2^4.5 further away, or around 1.8 miles.

This does not take in to ANY kind of diffration, fresenel zone effects, etc., etc.

HOWEVER, I can easily see how you could manage a link at least in the several score of Mbps with a more modern bridge pair at 3.5 miles (I've seen setups where people have managed >20Mbps at distances of >10 miles utilizing 802.11 still and all still being within FCC regulations).

It still may require some tweaking to the current setup, such as mounting the bridges closer to the antennas to reduce coax cable signal attenuation as well as configuration tweaks to the bridges to get them to behave well. You are also likely to see problems in bad weather. a few hundred feet through torential downpours or blizzards isn't likely to change the link strength much at all, even on 5GHz, but over 3.5 miles, atmospheric conditions, from rain to thermals is going to start impacting it quite a bit more, 2.4GHz, 5GHz, 3.6Ghz or 900MHz are all going to start getting skewed by the weather (which doesn't mean it won't work in bad weather, but an ideal day might allow 30Mbps of real through put on the link, but really bad weather might push it down to 5-10Mbps).

At any rate, greater than whatever the link rate was on the old bridges shouldn't be that difficult with some good newer bridges and tweaking the physical configuration as well as the bridge settings.

 

[stevech]

There are zillions of Cisco outdoor bridges in use with 26dBi small dishes in 5.8GHz. And other brands. I've engineered some of these point to point links - some are 5+ miles.

The big issue with those old Aironet bridges is they pre-date even 802.11b and are very low data rates. But that's good for range. I'd have used at least 12dBi antennas on each end, elevated to 25+ feet due to the path length.