Consideration of Nonlinear Noise and its Testing in Frequency Division Multiplex Voice UHF Radio Communication Systems

In UHF radio communications systems, the thermal noise limits the performance under weak received signal conditions. When the signal is strong, the nonlinear noise (sometimes called the intermodulation noise) may become the limiting factor. it is, therefore, imperative to control the intermodulation distortion contribution from the system components by careful equipment design and from the medium by appropriate choice of system parameters. In single-sideband systems, nonlinear noise may be attributed to two main causes: 1) transmitter nonlinearity and 2) receiver nonlinearity. Multipath effects in the medium do not seem to produce any nonlinear distortion, except selective amplitude fading or frequency distortion. Amplitude nonlinearity is the sole contributing factor in nonlinear noise. There is no set rule on the subdivision of nonlinear noise among both the transmitter and the receiver. Equal distribution seems to be a reasonable assumption. In some cases, an entirely different distribution may be desired, according to actual requirements. How to distribute the transmitter or receiver nonlinear noise among its components depends on the characteristics of the components. The general principle is to have the most economical design on every piece of the components, if possible and practicable. The nonlinear noise in single-sideband systems may be assumed to consist of only 3rd- and 5th-order products. It is further assumed that 3rd- and 5th-order products contribute to the total nonlinear noise in equal amounts in power. In frequency modulation systems, nonlinear noise may be attributed to three main causes: 1) transmitter nonlinearity, 2) multipath effect in the medium, and 3) receiver nonlinearity. Both amplitude and phase nonlinearities are equally important in the contribution to the noise and each should be considered carefully. The subdivision of nonlinear noise may be done in two ways. It may be distributed among the transmitter, the path and the receiver, or- among the amplitude and phase nonlinearities. In either way, further subdivision is necessary in order to specify the amplitude or phase linearity requirements for the various components of the system. There is, again, no set rule on the distribution of nonlinear noise among the various causes. The same principle as applied to singlesideband systems as mentioned before may be used for frequencymodulation systems. The nonlinear noise in frequency-modulation systems, may be assumed to consist of only 2nd- and 3rd-order products. It is further assumed that 2nd- and 3rd-order products contribute to the total nonlinear noise in equal amounts. Methods for performance evaluation, based on signal-to-nonlinear-noise ratio, are given and sample calculations are made to illustrate the methods. In the test of a complete system, either two-tone or noise loading tests may be used. Assuming that only 3rd-order nonlinear noise is predominant and that a peak factor of 10 db for random noise is chosen, the noise loading test signal-to-nonlinear ratio is 8 db higher than the two-tone test ratio if the rms power in the two tests are equal. However, if the peak power in the two tests is equal, both ratios are the same.