Multiple Invariance Spatio-Temporal Spectral Estimation (SpatiotemporalSpektralestimering med Multipelinvarianser)

摘要

In this work, the authors present an efficient high resolution signal subspacemethod for the simultaneous determination of center frequencies and directions-of-arrival (DOAs) of multiple narrow-band sources. The method is based on ESPRIT and an array with multiple spatio-temporal invariances. The authors show that by employing certain symmetries in the space-time sampling pattern the number of sensors can be kept very low relative to the number of signals, without relying on frequency separation via prefiltering. Moreover, these symmetries make it possible to solve the problem of pairing the frequencies and DOAs in a simple way. The 2-D parameter estimation problem is formulated in terms of three coupled 1-D problems that can be solved in parallel. Simulation results are provided to illustrate the application of the method.