I never saw any of the earlier messages from this thread, but I have a 4 sq mi mesh implementing right now. It is pretty much a tropical rain forest without the rain. I was required to use a dual radio product that only allows meshing on 5.8. The 2.4 is for client access only.
Due to the trees, and I don't even want to mention how many nodes/mile there are, it is very difficult to get signal injection into the outlying neighborhoods. I'm having to rely on VPN over cable modem as a last resort. However, my latest experiments using Ligowave under tree canopy (at the 21 ft height at which my nodes are mounted) have been very promising. Ligo is reasonably priced and offers a really decent penetration through the overhanging trees. I'm seeing -55 to -70 signals on my hoppity-hoppity-hop links. If I can do enough of it, I won't have to use any cable modems. The dual radio units are very nice because the Ethernet port can allow me to make the "host pole" a gateway without having to have a switch mounted there as well! Ligo on the pole, panels facing each direction up and down the street and plug the Ethernet into the mesh radio. I can concur with Brian. I was the one who had to be the 1 man SWAT team to go in and fix those Earthlink networks after they were implemented. Philadelphia was where I spent 80% of my time and it was an extremely difficult place to make things work due to the 4-5 story residences that were 6 feet from the street. I guess these are called brownstones. If you could not put radios at street corners, you were pretty much out of luck! One more thing to say to those unfamiliar with mesh. This goes with the paragraph two above this. A mesh quickly becomes so slow that it is useless without the proper number of gateways, or bandwidth injection points. Although you may be able to mesh all the way to the city limits (or at least out 6-8 hops), your speed will be a crawl. You have GOT to have that injection going on. Sometimes as small of a ratio as 3-4 mesh nodes per injection point. That is why if you drive through Anaheim, Philly, (and for another week- New Orleans), you'll see lots and lots of Canopy SMs hanging from the Tropos radios. That's the injection layer going back to a PtMP canopy cluster someplace.. Ralph -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Brian Webster Sent: Thursday, June 12, 2008 9:08 AM To: [EMAIL PROTECTED]; WISPA General List Subject: Re: [WISPA] using street level and below tree canopy for unlicensed bands In my days at EarthLink we did discover that the noise levels in both the 2.4 and 5 GHz bands were much lower at street level than up on high buildings or towers. This was both good and bad. It was good in that we had a better signal to noise ratio. The reason being that in Philly the buildings in many cases were three stories and taller than the mounting points we were allowed to use on the poles. With buildings shadowing a node it was much quieter. This was also a big problem in the network design. Those same buildings and trees also kept the mesh nodes from being able to link anywhere but straight along the streets between the buildings and even that was a challenge with the trees. For those reasons Philadelphia ended up with a much higher transmitter count per square mile than originally anticipated ( a lot more). Now the idea of shooting signal under the tree canopy is a good one. One problem is that you need to ensure that you can actually mount the radio below the tree canopy. In most cities the lower part of the canopy will be 10 to 15 feet above the ground and pole mounting heights typically are 20 feet or higher. At 5 GHz on a mesh system if you have to go through any more than about 10 meters of tree, the attenuation is such that you can't hold a link (with or without noise). If you are designing a network to shoot under the trees, you better have someone out on the field visually verify every single link that you want to work is clear of obstructions end to end. The reason being that any slight change in ground elevation can easily block the path because you have obstructions to consider both above and below your RF path. RF tools can not account for this, even if you have high resolution tree clutter data. They will model the tree as solid all the way down to the ground. They can not show the clear path area on the underside of the canopy. If this were a small mesh deployment and you could verify links on the ground I would say it was possible. Thank You, Brian Webster www.wirelessmapping.com <http://www.wirelessmapping.com> -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Rogelio Sent: Thursday, June 12, 2008 2:02 AM To: WISPA General List Subject: [WISPA] using street level and below tree canopy for unlicensedbands I was talking to some people today who deploy wireless networks in very noise environments, and some of them were talking about deploying radios under the building or tree line in an attempt to get less nodes. One person said that this practice is common in places like NYC where the street level is relatively free on the 5GHz band. Anyone else do or heard of this? (just curious...) ---------------------------------------------------------------------------- ---- WISPA Wants You! Join today! http://signup.wispa.org/ ---------------------------------------------------------------------------- ---- WISPA Wireless List: wireless@wispa.org Subscribe/Unsubscribe: http://lists.wispa.org/mailman/listinfo/wireless Archives: http://lists.wispa.org/pipermail/wireless/ ---------------------------------------------------------------------------- ---- WISPA Wants You! Join today! http://signup.wispa.org/ ---------------------------------------------------------------------------- ---- WISPA Wireless List: wireless@wispa.org Subscribe/Unsubscribe: http://lists.wispa.org/mailman/listinfo/wireless Archives: http://lists.wispa.org/pipermail/wireless/ -------------------------------------------------------------------------------- WISPA Wants You! Join today! http://signup.wispa.org/ -------------------------------------------------------------------------------- WISPA Wireless List: wireless@wispa.org Subscribe/Unsubscribe: http://lists.wispa.org/mailman/listinfo/wireless Archives: http://lists.wispa.org/pipermail/wireless/
