Comparing DSC and Voice Radio Sensitivity

Electrical and electronic topics for small boats
jimh
Posts: 3724
Joined: Fri Oct 09, 2015 12:25 pm
Location: Michigan, Lower Peninsula
Contact:

Comparing DSC and Voice Radio Sensitivity

Postby jimh » Mon Sep 11, 2017 9:16 am

Specifications for sensitivity for a radio generally give the radio signal level at the receiver input which will produce the demodulated signal with a certain fidelity or accuracy at the receiver output. The nature of the specified accuracy or fidelity varies with the type of modulated signal.

In the case of voice modulated transmissions, the sensitivity is generally specified in terms of the signal to noise ratio of the output (loudspeaker) for a given radio input signal. Since it is difficult to measure the noise without including any distortion in the actual signal, the measurement is usually specified as signal to noise and distortion. This ratio is expressed in decibels and called the SINAD ratio. The measurement is typically made with a single audio frequency tone as the modulation which modulates the carrier 100-percent. A tone of 400-Hz is often used. When a voice receiver is specified as having a sensitivity of, say, 1.0-µVolt, for 12-dB SINAD, it means that the receiver can recover the modulation from an input signal of that level and produce an output signal with the specified signal to noise and distortion.

In digital radio systems, the same general technique of specifying a particular input signal level which produces a demodulated output of a specific fidelity is also used. The radio signal is specified in a similar manner by giving its level, either in micro-volts or in decibels relative to one micro-volt (or dBµV). The modulation in a digital radio is digital data, so there is no particular usefulness is specifying a signal to noise ratio for the data, nor is the output a loudspeaker. Instead the fidelity of the demodulated signal is specified by its accuracy to the original. The rate at which an error in the digital data may occur is usually given as a numeric probability.

For an example, we can look at a recent VHF Marine Band DSC radio. The manufacturer specifies as follows: the communication radio is rated -5 dBμ emf (or 0.28 μVolt) for a 20-dB signal-to-noise-and-distortion (SINAD) ratio, and the AIS receiver is rated -3 dBμ emf (or 0.35 μVolt) for a one-percent bit error rate (BER).

We see that the radio sensitivity is about the same for both modulation methods. In the case of voice transmission, the information in the modulation is extracted by the listener and his aural perception. In the case of DSC the information is extracted by a digital demodulator. In both cases the performance can vary, either by the listening skills of the individual or the circuit design. There are many levels among humans in their ability to copy voice radio signals in the presence of noise and interference. Similarly, not every circuit designed to demodulate a digital signal will perform uniformly.

One element that favors the DSC digital signal: very bad modulation on the transmitter on voice, due to terrible microphone techniques. I hear many VHF Marine Band voice transmissions where the modulation level is so low that the modulation is barely audible, even though there is excellent carrier level and very high noise suppression.

The duration of a DSC transmission is quite short, less than a second, and conveying the same data by voice could take much longer, perhaps a minute. In the case of DSC transmissions, the protocol provides for really important messages, such as a distress alert, to be retransmitted several times in the short burst, giving some redundancy, and to retransmit the entire message automatically at intervals, until acknowledged. To get good reception in an environment with fading signals requires only a brief interval of good signal with DSC, while with voice and a longer transmission, a signal could completely fade out at some part of the transmission and that data would be lost. Computing the statistical probability of duration of strong signals and the coincidence of a short transmission length in the presence of random signal fading would seem to be a random number.

A further advantage to DSC is the demodulator is always paying attention, while a human listener may not be so diligent.

jimh
Posts: 3724
Joined: Fri Oct 09, 2015 12:25 pm
Location: Michigan, Lower Peninsula
Contact:

Re: Comparing DSC and Voice Radio Sensitivity

Postby jimh » Mon Sep 11, 2017 9:29 am

For a further discussion regarding voice signals and AIS signals, which, like DSC, are digital modulation, see my earlier remarks at

Comparing Range for VOICE and AIS Transmissions
http://continuouswave.com/ubb/Forum6/HTML/003657.html