A Monte Carlo method for predicting peak values of random sums

Prediction of a reasonable value for the sum of randomly varying signals of different modulation forms and different power levels by a Monte Carlo procedure is discussed. Random signals are defined here as those signals with amplitude variations which, according to the rigorous definition, may not be truly random but have amplitude variations that are too complex to be easily described mathematically. The Monte Carlo method gives an analytical insight into the nature of signal loading of a multichannel power amplifier. The effect of the crest factor of modulated signals is demonstrated and seen to be a significant factor affecting the loading. Frequency modulation (FM), phase-shift keying (PSK), and other sustained carrier forms of modulation consume a dominant amount of the amplifier´s dynamic range and thus are significant in establishing signal loading criteria. As the number of signal components increase, the crest factor of the composite tends to converge to a limiting value. As few as five components will converge the composite to within approximately 1 dB of the limiting value