Design factors for the optimization of multichannel radio systems

The satisfactory design of a multichannel radio system requires evaluation of both nonlinear and thermal noise power and the statistical considerations associated with each. It is pointed out that an optimum design can be attained when these noise powers are equal; and since the noise power is greatest in the top channel of the FM (frequency modulation) system, it is sufficient and suitable to establish design criteria based on these premises. It is seen that certain essential performance factors are required for this purpose. These include a measurement of signal/noise ratio using a white-noise test signal, a knowledge of the principal orders of distortion which give rise to the nonlinear noise, and a value for the overload power of the base-band signal. The effect of such factors as carrier power, channel spacing, speech-power distribution, and traffic activity, on the signal/noise ratio is then examined. The basic expressions are also used to establish the total reliability as a function of traffic fluctuations, propagation reliability, and carrier power. Theoretically optimum distributions as well as typically operative distributions are given. In general, systems are of two classes: integrated ¿ those which must conform to the technical requirements of existing or proposed associated networks, and non-integrated systems ¿ those which exist, as it were, independently. Standards are suggested for both these systems. The CCIF (Comite Consultatif International Telephonique) fictitious circuit of reference is a special case of the former and it is concluded that the signal/noise ratio suggested for this reference is some 10 db (decibels) too high. System testing by means of multiple tones is also discussed and some basic requirements related to this method are outlined.