Measuring radio frequency interference is key to avoiding it

Certain interference can limit the ability of a base station, mobile or portable to capture desired signals. This phenomenon is termed "effective receiver sensitivity" and is caused by in-band noise that raises the noise floor and minimizes the ability to listen to the desired channel. If a receiver has a sensitivity of -125 to -130 dBm on the bench, it may deteriorate to -115 dB in the presence of a disruptive noise environment. This deterioration of the receiver's performance limits the range in which a mobile or portable radio can talk back. Where a mobile or portable has a talkback range of 20 miles with no external noise, range may be cut in half with the presence of 10 or 15 dB of noise.

Intermodulation also can create the same effects as external noise. As transmitter signals mix and create new unwanted signals, the desired signals can be muted or distorted — again, the effect is deterioration of the base station's range, as well as the range of the receivers, at the site. This interference is much more difficult to detect and mitigate, because the effects are mostly intermittent. For voice systems, the intermodulation manifests as loud sounds when the signals are analog in nature; in digital systems, the speech will drop entirely. In both cases, signal quality and range are negatively affected.

Unwanted signals close to the desired signals also can cause receiver desense that is similar to the first two types of interference and with similar effects — i.e., either a reduction of range or audibility.

The effects of noise, intermodulation and desense can be modeled in a radio-frequency link budget or an RF tool using C/N; C/I; C/(I+N); C/A; or C/(I+A+N). Let's examine each.

Carrier to noise, or C/N, represents the amount the desired signal must be above the effective receiver noise floor in order to operate efficiently, and is displayed in dB. A typical value for C/N would be 7.6 dB for Project 25 Phase 1 radio systems.

Note that signal fades and required fade margins vary between fixed, nomadic and fast-moving subscribers in an RF environment. For example, a fixed subscriber experiences slow, or Log-Normal, fading due to the nature of the radio wave fronts and how multipath reacts and is received. Log-Normal fades are weak, in the range of 3-6 dB for 95% confidence, and are long-lived. Because the receiver is fixed and cannot move out of the fade, these values should be considered to have a high probability of occurrence. If antenna spatial diversity is used at the base-station site or at the subscriber end, the value can be corrected by 3-6 dB. Another optimization technique called MIMO (multiple input, multiple output) leverages multiple antennas to reduce the impact of this type of fade.