I need to make a 7 mile wifi link that will stand up to interference like snow. Speed is not an issue. I am fine with only 1Mbps. I was thinking about using 24dbi dishes on each end, and radios with TX
18dbm, and a sensitivity of -80dbm. I will mount the radios in water tight enclosures next to the dishes.
Well, let's do the math using your numbers. Well, maybe not. They're wrong and need work. The typical receiver sensitivity at 1Mbit/sec is
-89dBm, not -80dBm. That's a huge difference.
Most bridges manage to squeeze out +15dBm, not +18dBm (despite what the data sheet claims). Just to add some confusion, the power output is usually slightly higher for 802.11g than for 802.11b.
You also don't wanna run 1Mbit/sec because it hogs way too much airtime. Methinks the best for a bridge is the slowest ODFM speed of
6Mbit/sec. The sensitivity is roughly the same as 1Mbits/sec. It should give you 3Mbits/sec thruput, which methinks is good enough.
The object of this exercise is to get a fade margin of greater than
20dB. Any less will result in way too much dropouts and errors.
TX Power : +15dBm TX Coax Loss : -4dBm TX Antenna gain : +24dBm Path Loss : (7 miles) RX Antenna gain : +24dBm RX Coax Loss : -4dBm RX sensitivity : -88dbm
yields a fade margin of 21.9dB. Yep, that should work. I would like to see more than 20dB fade margin, but I don't think that's going to happen without more tx power.
I'm not sure how much snow will affect the path. Probably not much except where it detunes the antenna. You might consider an overpriced radome to fit over the dish.
Craig discusses some of the totgraphical issues, and David Taylor has an interesting story. If the 7 mile link doesn't work, maybe you could do a hop in the middle.
formatting link
documents a 7KM link. Oops, KM, not M... still, the process works.
See if you can find an old copy of CCIR XIII Plenary Assembly in Geneva,
1974, Volume V, Report 233-3. This one is often referenced in stuff on propagation loss studies. See the Electromagnetic-Wave Propagation chapter in the ITT Reference Data for Engineers (from Sams) as an example. On the _path loss_ alone, snow is not a factor below 6 GHz.
From a paper produced for NASA Ames in 1984:
----------- One might ask if the half inch per hour rainfall rate is realistic. In some regions of the country, this would be an overkill, while in some other areas, it is not nearly enough. Other forms of precipitation cause different effects. Snowfall at a rate of 15 centimeters (5.9 inches) per hour with a liquid equivalent of 15 millimeters per hour has little effect up to 10 GHz. At 18 GHz, attenuation is approximately 0.01 decibels/kilometer. Hail has an attenuation effect dependent on the state. Dry hail is generally small, compared to raindrops of the same liquid content. As the hailstone (or snowflake) begins to melt, the attenuation increases to the point (with a diameter of a tenth wavelength, and 10 to 20 percent melted) it may be double that of the equivalent liquid content. Fog (or clouds) produce attenuation based on visibility and frequency. Table 3 lists typical attenuation rates for various visibilities.
Depends on the weather, but MAY be a good idea. Don't forget to include a slight increase in the path loss due to the radome, and the change in wind loading on the supporting structure..
Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.