l Clearing Fresnel Zone: Irrespective of whether the radio link planned is point-to-point or point-to-multipoint, the first thing to do is to verify that it will have not only clear line of sight, but at least 60 percent of the first Fresnel zone clear of obstructions as well. The longer the distance, the more important it is for this verification. If the Fresnel zone is blocked, then you will get a lower signal level on the distant end than expected—even if you can literally ‘see’ the other antenna in the distance.

l Perform RF-path Analysis: Even if the Fresnel zone is partially blocked, it is still possible to get a link, provided that the radio system is designed to have a strong signal at the other end of the link. In planning long-range microwave links where unobstructed line-of-sight and clear Fresnel zone are not certain, an RF path analysis should be done. There are many software packages available that have terrain data and can create a path profile from a set of latitude/longitude coordinates.

l Have Unobstructed line of Sight: The software programs for RF path analysis can only indicate for certain if a link will not work due to terrain obstruction. A clear path on paper is not a guarantee that your link will work, since it does not show trees or buildings. So, even if there is a ‘clear’ link on paper, RF analysis may show if there are 80-feet trees that can block the signal.

l Perform SOM Calculation: In case of both clear line of sight and 60 percent of the first Fresnel zone clear (or nearly clear), how can you know if you will have a good link or not? How much gain do your antennas need to have? How much coax cable loss is too much? If your link is at 2.4 GHz, should external amplifiers be used?

Or, given your fixed base station antenna with a pre-set gain, how far can you reach with the different types of client antennas? And, which clients will need amplification? By doing an SOM calculation, you can test various system designs and scenarios to see how much fade margin (or "safety cushion") your link will theoretically have.

The SOM is determined by computing the difference between the received signal and the radio’s receiver sensitivity. The typical formula used is

RX Signal = EIRP – FSL + RX Antenna Gain – Coax Cable Loss

Regarding the minimum SOM needed, there is no absolute answer to this question, but the higher it is, the better. Most engineers agree that 20 dB or more is quite adequate. Some think as low as 14 dB is still good. Others operate systems down to 10 dB or less.

The problem with accepting a lower SOM is that you have a smaller safety margin. You run the risk of your link going down in case of interference, an antenna off its aim, atmospheric conditions, moisture in your coax, ice/rainwater on the radome, or a host of other factors.

l Determining Interference: The SOM is not the only determining factor. It is the actual SNR at the receiver that makes a link reliable. If there is noise or interference on the channel, the SNR will deteriorate. This could be an issue if you are co-locating at a site with other radios operating in the same band.

You need to find out what frequency spectrum these radios are occupying. If these transmitter have energy or sideband noise on your receive channel and their antennas are close to yours, you will likely get interference from them, perhaps to the point where your link will not work.

l Factoring Atmospheric Absorption: The SOM calculation holds perfectly for a vacuum. In reality, there is some atmospheric absorption of the RF energy that scatters and attenuates the signal. For example, tests on a 23-mile 5.8 GHz link vary as much as +/-6 dB over course of a day. This variation is mostly caused by multipath interference and other atmospheric variations.

l Using an Amplifier: With coax cable at the receiver and no amplifier at the receiver antenna, the SNR at the antenna does not survive when it actually reaches the radio itself. In this case, the noise generated in the RF front-end of the radio is a factor.

If an amplifier is used on the receive end, the SNR as it appears at the antenna is preserved all the way down the co-ax to the radio. This phenomenon largely occurs because the low-noise amplifier mounted on the pole sets the noise floor for the system.