Necessary conditions for estimating the maximum bit error probability by using amplitude probability distribution of an interfering noise

Abstract

Necessary conditions of a digital modulation signal are investigated in order to estimate its maximum bit error probability (BEP) by using the amplitude probability distribution (APD) of an interfering noise. It was found that the waveform difference between any two symbols needs to be able to be normalized to have an identical APD and amplitude spectrum. Moreover, the spectrum needs to match the frequency selectivity for the APD measurement with a tolerable deviation. Some modulation schemes are discussed with respect to the conditions.