In order to study the effects of noise upon radio communication, an amplitude-modulation radio system was set up in the laboratory and provision was made for generating electrical interference and introducing it, together with the speech-modulated carrier from the transmitter, into the receiver. In addition, arrangements were made for producing ambient noise at the talkers' and listeners' positions, so that the separate effects of acoustic noise and radio noise could be compared and their combined effects could be studied. By means of word articulation tests, the intelligibility of speech heard over the radio system was determined as a function of signal-to-noise ratio for each of a number of types of noise. Several principles of noise reduction were studied, and basic parameters of the radio link were varied systematically so that their influence upon the effects of noise could be determined. With regard to ambient noise, it was found that exclusion of noise at the microphone is even more important as a prerequisite for effective radio communication than it is for effective interphone communication, especially if either compression or premodulation clipping is employed in the radio transmitter. Noise exclusion at the listeners' end of the line is also important because, for optimal intelligibility under difficult conditions, it is necessary for the noise reaching the ear through or under the earphone cushions to be at least 10 decibels less intense than the noise coming through the earphones from the receiver. The deleterious effect of electrical interference was found to depend greatly upon the relation between certain characteristics of the noise and corresponding characteristics of the receiver circuits. In general, interferences with continuous spectra are more detrimental than those with line spectra, and non-impulsive types more detrimental than impulsive types. Noise-reducing circuits were ineffective against random fluctuation noise, but in the presence of certain types of impulse interference, limiters and cancelling circuits provided such great improvement in performance that it was possible to maintain satisfactory communication despite a 35-decibel reduction in carrier intensity. As a general principle, it appears that, whenever there is a characteristic difference between the wave forms or the spectra of the signal and the interference, the impairment of intelligibility by electrical interference may be reduced by employing amplitude-selective or frequency-selective circuits in the radio receiver.
Read full abstract
Sep 1, 1948
D Weighton 