Abstract: The order statistic signal processor obtains continuing estimates of amplitude, phase, or frequency of an input signal based on order statistics. By way of definition, the P'th order statistic for a set of samples from a continuously distributed process is the P'th largest of the samples. Where a signal parameter fluctuates with time as a result of contamination by noise and interference and where the noise fluctuation rate is much greater than the rate at which changes in a signal parameter occur, an order statistic associated with a set of signal parameter samples can serve as an estimate of the signal parameter. The order statistic signal processor is based on an iterative process whereby the prior estimate of a signal parameter, i.e. amplitude or phase, is subtracted from each of a plurality of signal parameter samples obtained over a period of time. The differences are converted to quantities of fixed magnitude with signs corresponding to the differences.