Receiver Sensitivity and Signal Levels

I keep hearing that I need a more sensitive receiver for my wireless system to work optimally. I understand this is related to decibels (dB) and signal-to-noise ratios (SNRs), but I'm not sure how. Will you please tell me how all this relates to signal levels?
Signed, Sensitive

Wise Guy: Received signal strength is not the sole indicator of communications system performance. For example, the received signal may be very high, but if the noise level is higher the bit error rate will increase and the channel will perform poorly. Conversely, the received signal level may be low, but if the noise level is even lower then the channel, having a low bit error rate may perform quite well.

 Wise Guy
Wise Guy

A better metric is signal-to-noise ratio (SNR), which is mathematically defined this way:
SNR (dB) = 10 log10 (signal level/noise level)

Figure 1 provides a conversion table for signal levels in watts (W) and in decibels relative to a milliwatt (dBm). (Note that dB is a unitless, or absolute, numerical value whereas dBm sets the reference point such that 0 dBm = 1 mW.) To discover SNR using this table, determine your signal level (in whatever units you want) and noise level (in the same units), and then divide the signal level by the noise level. Note that SNR is a number (no units because the signal "units" were cancelled by the noise "units").

Figure 1. Comparison of different scales for power levels
Figure 1. Comparison of different scales for power levels

While a more sensitive receiver is, in fact, more sensitive, the SNR dictates how well the communications channel will perform. A standard rule of thumb is that you need an SNR of 10 to get the show started. But as they say in automobile ads, your mileage may differ.

The Effect of Water on Signal Propagation

Why doesn't my wireless signal propagate for a longer distance underwater? And how does humidity in the atmosphere affect above-water signals?

Signed, Drenched and Clueless

Wise Guy: Water absorbs certain frequencies and, based on this absorption, the amount of water present within the atmosphere (as rain, humidity, clouds, etc.) causes variation in the transmitted signal.

Figure 2, a graph redrawn from a 1980s U.S. Navy document (CCIR Report 719, 721), shows the effect of atmospheric water on signals. Note that it depicts signals in terms of wavelength, not frequency. Engineers tend to think in terms of frequency, while chemists and physicists tend to think in terms of wavelength.

Figure 2. Attenuation effects due to water in the atmosphere
Figure 2. Attenuation effects due to water in the atmosphere

Here, blue lines depict signals affected by rain and green lines depict signals in fog, while the red line indicates a signal traveling through a dry atmosphere. Note that the shorter the wavelength (higher frequency), the more the signal is absorbed.

Wise Guy is the problem-solving persona of WINA, the Wireless Industrial Networking Alliance (www.wina.org). Send your questions to [email protected].