Multi-domain adaptive parameter estimation

The present study is concerned with the problem of obtaining optimal estimates of several signal parameters each subject to inter-related constraints. The estimates are obtained by adaptively processing a set of observations or measurements which may be taken from any one of several possible observation domains. According to the employed approach, the parameter estimates are obtained from a set of weight vectors which operate in a linear manner on a vector of observations or measurements. The weight vectors are chosen to yield the desired signal parameters in the absence of noise or interference and to yield minimum variance estimates in the presence of background noise for which the statistical properties are specified. The relationships are illustrated with the aid of an example from the radar/sonar field in which the signal is defined in terms of its waveform in time and the shape of its wave front in space.