A tutorial review of the new EMI (Electromagnetic Interference) models and their effects on receiver performance

Models of electromagnetic interference (EMI) developed over the last five years (1974 to 1979), for the first time provide canonical, analytically tractable, and experimentally well established quantitative descriptions of nearly all EMI environments. These models are: (1) physically derived; (2) canonical in the sense that they are invariant of the nature and waveform of the source and details of preparation, as far as their formal analytical structure is concerned; (3) are highly nonGaussian; and (4) are analytically and computationally manageable. Their principal quantitative and most widely applied form is embodied in the first order probability distributions of the (instantaneous) amplitude and envelope of the received waveform following the linear front end states of a typical receiver. Three basic EMI models are distinguished: class A, B, and C, respectively involving sets of three, six, and eight, physically derived parameters which are measurable from observed EMI amplitude (or envelope) data. These three basic classes are defined in terms of receiver bandwidth vis-a-vis that of the EMI.