Exact Phase Retrieval in Principal Shift-Invariant Spaces
Data(s) |
2016
|
---|---|
Resumo |
We address the problem of phase retrieval from Fourier transform magnitude spectrum for continuous-time signals that lie in a shift-invariant space spanned by integer shifts of a generator kernel. The phase retrieval problem for such signals is formulated as one of reconstructing the combining coefficients in the shift-invariant basis expansion. We develop sufficient conditions on the coefficients and the bases to guarantee exact phase retrieval, by which we mean reconstruction up to a global phase factor. We present a new class of discrete-domain signals that are not necessarily minimum-phase, but allow for exact phase retrieval from their Fourier magnitude spectra. We also establish Hilbert transform relations between log-magnitude and phase spectra for this class of discrete signals. It turns out that the corresponding continuous-domain counterparts need not satisfy a Hilbert transform relation; notwithstanding, the continuous-domain signals can be reconstructed from their Fourier magnitude spectra. We validate the reconstruction guarantees through simulations for some important classes of signals such as bandlimited signals and piecewise-smooth signals. We also present an application of the proposed phase retrieval technique for artifact-free signal reconstruction in frequency-domain optical-coherence tomography (FDOCT). |
Formato |
application/pdf |
Identificador |
http://eprints.iisc.ernet.in/53167/1/IEEE_Tra_on_Sig_Pro_64-2_406_2016.pdf Shenoy, Basty Ajay and Mulleti, Satish and Seelamantula, Chandra Sekhar (2016) Exact Phase Retrieval in Principal Shift-Invariant Spaces. In: IEEE TRANSACTIONS ON SIGNAL PROCESSING, 64 (2). pp. 406-416. |
Publicador |
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC |
Relação |
http://dx.doi.org/10.1109/TSP.2015.2481871 http://eprints.iisc.ernet.in/53167/ |
Palavras-Chave | #Electrical Engineering |
Tipo |
Journal Article PeerReviewed |