Random asynchronous sampling strategy for measurement instruments based on nonlinear signal conversion

A random asynchronous sampling strategy is proposed for the digital implementation of broadband measurement instruments which are based on nonlinear signal conversion. A theoretical analysis of the basic features of this method is given by considering discrete-spectrum almost periodic signals. It is shown that, even when the band of the signals extends beyond the average sampling frequency, this sampling strategy does not introduce systematic errors in the measure, but only random noise which can be adequately filtered when the time available for the measurement is long enough. A formula for the evaluation of the standard deviation of the measure in terms of the spectral characteristics of the signals is given. Some of the specific problems related to the practical implementation of the sampling strategy are also discussed. As a preliminary validation of the technique, the results obtained by simulating the proposed instrument are compared with the theoretical error analysis.<>