FIELD: communications.
SUBSTANCE: means for exchanging voice messages. Signal processing is carried out in two channels, the main (first) and the additional (second). Preliminarily, in both channels in the absence of interference, samples of the speech signal are taken, in the interval that does not contain a speech signal, interference samples are taken in each channel. The average values of the amplitudes of the signal and interference are calculated, then the coefficient of change in the amplitude of the interference and the change in the signal in the second channel relative to the first channel is calculated. After appearance of the speech signal, the amplitudes of the mixture of interference and signal are measured and the value of the signal amplitude is calculated using the amplitudes of the mixture of interference and signal in each channel, as well as the values of the coefficient of change in the amplitude of the speech signal in the second channel relative to the first channel and the value of the coefficient of change in the amplitude of the interference in the second channel relative to the first channel. In pauses between transmissions of voice messages, which, among other things, are organized when the quality of received speech information deteriorates, interference samples are taken in each channel, the value of the interference amplitude change coefficient in the second channel relative to the first channel is specified.
EFFECT: ensuring selection of a speech signal from acoustic noise.
1 cl, 2 dwg
