I would like to note that currently, this bit of information is just a copy + paste of some old Facebook posts. Given time, this document will be massaged into something a bit more cohesive. The intent at the moment was just to get the information out and into something easier to find.
ITU Rain Zones
Do you have doubts about the abilities of the ITU prediction models to accurately reflect the uptime performance on higher frequency links? I sure do. Averaging rain out over the course of a year isn’t producing models that accurately reflect the high intensity of storms that we’ve been getting the past few years. Here is perhaps a better way of calculating if your higher frequency link will make it or not.
Calculate your system operating margin based upon normal conditions. You should end up with a signal level somewhere between -20 and -60. Hopefully more signal rather than less. Determine how much signal you can lose before losing full modulation and before losing the link altogether. The difference between what you should get in the best conditions and these two levels are your fade margin. Now determine the amount of rain fade in your frequency. This would be how many dB per kilometer you will lose based upon a rain rate of 1 inch per hour. Take that rate as appropriate for the distance of your link. Double it if your link is 2 kilometers, while you cut it in half if your link is half a kilometer. With that new number which would be decibels of loss on your particular link per inch per hour of rain, you can determine how many inches per hour of rain you can take on your link before you lose full modulation and before you lose the link altogether.
What good is that? Take a look at http://dipper.nws.noaa.gov/hdsc/pfds/pfds_map_cont.html?bkmrk=il and choose a station near your link. Change the drop-down at the top to say precipitation intensity. Now the top row of the resulting table should tell you how many inches per hour with 90% tile accuracy you are likely to get in your area when looking at a 5 minute average over a number of years. Take the inches per hour that you came up with in the last paragraph for how much you need to break full modulation and to break your link entirely as well as find the number of years you are likely to be able to go without encountering those numbers. That should give you a much better indication as to the reliability of your link.
24 GHz vs. 80 GHz
Running some numbers for 24 GHz and 80 GHz, for the link distances they’re considered reliable at around here (Chicago area), it looks like 24 GHz and 80 GHz have similar reliability over links in the 1-mile area, only 80 GHz will have more capacity due to larger channels.
80 GHz loses somewhere around 15 – 20 dB/km more in a good rain than 24 GHz, but 80 GHz has higher power levels, so your link is 20 – 30 dB hotter.
Also, to get comparable throughput, 24 GHz uses a more advanced modulation, which means it will be more fragile in the rain than 80 GHz, which will be using a more simple modulation due to drastically larger channels.
Neither are really to be considered reliable beyond 2 miles here.
No doubt the 80 GHz gear costs more than the 24 GHz gear.
The upper band 60 GHz gear should have similar RF characteristics to 80 GHz gear, but should also be quite a bit less expensive.
Please comment on this Google Sheet Mike put together. It’s basically taking his rain fade text wall from two years ago and putting it into a calculator.
You put in frequency, distance, transmit, receive, etc. parameters and it helps you determine if you can afford that link in terms of rain fade.