Slashdot Mirror


Building an Non-Wired Network for Pueblos?

wsxian asks: "We recently received a large grant to install a wireless high speed Internet system for our entire Pueblo here in the State of NM. This would encompass about 150 homes and the range from a central point would be no more than 6 miles, but there are hills, valleys and trees so line of sight is not an option unless we decide to drop homes. I welcome any suggestions for technology to look at, what to avoid, security concerns- and most important for us what companies to avoid. If any of you have had real life experience with doing such a scaled project like this, please give your input!"

7 of 60 comments (clear)

  1. No Non-Line-of-Sight by Neon+Spiral+Injector · · Score: 4, Informative

    No matter what anyone tells you, or tries to sell you: wireless is line-of-sight. Microwaves are electromagnetic radation. They do penetrate a bit better than the visible wavelenghts, but a building or a hill, heck even trees, will put a real cramp on your transmissions.

    If all buildings don't have a clear, unobstructed view of a central point; the only way to make this work is to connect the buildings that can see eachother until you reach one that can see your point of origin.

    Start by finding the best location that is the most visible, and most central to the most buildings. This is were you will install six radios each with a 60 degree sector antenna on a tower (less if you don't need a full 360 coverage, but don't go any wider then 60 degrees for each sector, or you'll be spreading your power too thin). Then survey each of the other locations to see which have perfect views of this tower. These locations will get an uplink radio with a unidirectional antenna focused on the tower. Plus a second radio operating in a different frequency range with an omnidirectional antenna. Try to use a lower frequency for the omni radios. These secondary radios will operate as a mesh network to pick up the other buildings that don't have line-of-sight to the central point. Finally the non-line-of-sight meshed-in locations can operate on their own omnidirectional antenna if they are close enough so they can pick up tertiary meshes. You really don't want to go any deeper than three into the mesh, speed starts to suffer too much. If the buildings are more than a couple hundred feet from each other the powerloss from transmitting in all directions will be too much and you'll have to go to directionalized antanna at one or both locations to direct the RF in a more focused pattern.

    1. Re:No Non-Line-of-Sight by VeriTea · · Score: 5, Informative
      I design wirless networks for a living, so I'd like to add a few words...

      Get a good set of topo maps for the entire area you want to cover. Don't let anybody fool you, real paper maps and a good distance calculator like "Topo Companion" are the best way to do wireless design. Next, go to the FCC and download their Antenna Registration Database. You will need the "EN" and "CO" and one other table. Join the tables (useing MS Access would be easiest) using the unique site id and then query out the sites that have coordinates within the boundaries of Pueblo (latitude > y1 and latitude x1 and longitue Now use your "Topo" companion to mark existing towers on the map. Look for ones that are on hills or mountains near the desired coverage area. Some simple trig will tell you if you have line-of-site (LOS). Add the average tree height to your ground elevation when doing calcuations, 802.11x is very low power and high frequency, so you can't afford any pathloss burning through vegetation.

      Unlike the parent, I would strongly recommend 30 degree panels for this kind of system, since you are likely to encounter capacity issues, and even log-periodic 60 degree panels have enough overlap between sectors to guarentee a huge interference problem if you use more then one transmitter. Find the equipment you are going to use and use your gains and losses to calculate the ERP / EIRP from your transmit antennas at each site. Now go ahead and use the free-space pathloss equation to figure out the signal strenght at you target locations. If you have adequate signal to meet the needs of the subscriber equipment you plan on using then great (Note: if you plan on supplying subscriber equipment then you can guarentee a minimum quality. If not, your design will need to be especially robust to accomodate low-quality receivers). If not, keep tweeking the design until it will work. Develop a plan for directing the different channels away from each other. Make sure you calculate your expected interference from other sites using the same channel. If you several mountaintop sites this will be a killer.

      Once you have it built on paper, the rest is fairly straightforward. Have fun!

      --
      --- There are two kinds of people, those who accept dogmas and know it, and those who accept dogmas and don't know it
    2. Re:No Non-Line-of-Sight by Neon+Spiral+Injector · · Score: 3, Informative

      About the 60 degree sectors. You are correct about the possiblity of interference. The radio equipment that the company for which I work uses each tie into a central control unit that provides a timing signal to each unit to stagger transmissions and prevent problems.

  2. Vivato outdoor wireless by GlL · · Score: 5, Informative

    Check out http://www.vivato.com/. I have been installing their products, and recently did a baseball stadium. The signal penetrated the concrete construction into the team offices behind the dugout. Good stuff.

    --
    I'm a happy pessimist. I expect and prepare for the worst, when it doesn't happen I am pleasantly surprised.
  3. Re:amen! by Joe5678 · · Score: 2, Informative

    Grants are not always from the government...

    A grant is simply a sum of money that has to be used for a specific purpose; and in this case, I would say almost certainly, wasn't from the government.

  4. BPL: Broadband over Power Line by choochus · · Score: 3, Informative

    I live in Manassas, Virginia which is the first US city to deploy BPL technology. The city has partnered with a company called COMTek, Inc. to supply the equipment and run their billing.

    I went to a demonstration with City officials and COMTek engineers and it was VERY impressive. Their technology will interoperate with any other TCP/IP transport (cellular radio, fiber, copper, etc.) and, in fact, Manassas uses fiber for their long haul.

    Comtek said that you can currently get 4 megabit to end users and with repeaters can push about six miles. At least, I think it was six miles, it's been a little while. Buy anyway, you can use radio to bridge long line of sight gaps, then stick a BPL inducer on a nearby transformer and light up any house within a mile of it, easily.

    One thing they did in downtown Manassas was light up a building w/ BPL and put a wifi hot-spot on the roof which gives downtown strollers high speed wireless Internet (as long as you are a subscriber :o)

    The technology is very flexible and very stable in a wide range of temperatures which is ideal for New Mexico. The BPL "modems" (for lack of a better term) cost around $200 a piece retail and the inducers can be installed on power transformers in about 30 minutes, so it is a very rapid deployment.

    Oh, and COMTek had installed some hardware that looked for amateur radio operators and dropped pieces of their spectrum so that they wouldn't interfere with them. I thought that was pretty cool of them as it does slightly lower their overall capacity, but not by much.

    Good luck!

  5. 900 MHz by LoaTao · · Score: 2, Informative

    I work for a WISP and would suggest that you look into 900 MHz for your system. 900 MHz is not pure line of sight and gives you about a five mile range from tower to end user modem to work with. We have WaveRider equipment deployed (www.waverider.com) and have had good results from it. Internal and external antenna installations are both options.

    One thing I would recommend is having an experienced professional do the network design. It's one thing to slap a 802.11x WAP into a building and plug it into the DSL modem and something very, very different to design a wireless WAN that will actually work.

    --
    The smartest man in the whole, wide world really don't know that much. - Mose Allison