Space-Frequency Decomposition and Time-Reversal Imaging

This paper presents two novel time-reversal algorithms: Space-frequency DORT (SF-DORT) and SF-multiple signal classification (SF-MUSIC) to image targets based on the SF multistatic scattering data matrix. Singular value decomposition is applied to this matrix to obtain the left unitary matrix, the vectors of which are classified into the signal-subspace vectors corresponding to the targets and the noise-subspace ones. Numerical simulations show that the imaging performances of the SF-MUSIC and the SF-DORT are similar to those of the time-reversal MUSIC (TR-MUSIC) and the time-domain DORT (TD-DORT), respectively. However, the new algorithms require only one single SF measurement by the TR array, in contrast to N measurements needed in the TR-MUSIC and the TD-DORT, when the TR array has N antennas.