Improved Microwave Imaging by Wavenumber Domain Multiband Data Fusion

Research output: Chapter in Book/Conference proceedings/Edited volumeConference contributionScientificpeer-review

10 Downloads (Pure)


A matrix-pencil based wavenumber-domain multiband fusion approach is proposed for improved microwave imaging. The suggested fusion approach is based on the fact from the Born approximation that the wavenumber-domain signal in a given scattering direction can be represented as a sum of the same number of contributions over the whole bandwidth. Utilizing this fact, the wavenumber-domain signal in each radial direction can be modeled as a superposition of damped/undamped exponential functions, where measured multiband data are part of the observations of the whole-band signal. Then the multiband signal fusion is addressed by estimating a unified signal model over the while bandwidth with the matrix pencil approach in the wavenumber domain. By using the estimated signal model, the missing data in the frequency gap can be extrapolated, thus synthesizing an equivalent wideband signal spectrum. After an inverse Fourier transform, the synthesize spectrum leads to a focused image with enhanced resolution. Compared to the existing frequency-domain method, the proposed fusion approach can be used for radar imaging with the signals acquired by either collocated or non-collocated arrays in different frequency bands. The effectiveness and accuracy of the proposed fusion approach are demonstrated through some numerical simulations.
Original languageEnglish
Title of host publication2018 International Conference on Radar, RADAR 2018
Place of PublicationPiscataway, NJ
Number of pages6
ISBN (Electronic)978-1-5386-7217-4
Publication statusPublished - 2018
EventRADAR 2018: International Conference on Radar 2018 - Brisbane Convention and Exhibition Centre, Brisbane, Australia
Duration: 27 Aug 201830 Aug 2018


ConferenceRADAR 2018
Internet address

Bibliographical note



Dive into the research topics of 'Improved Microwave Imaging by Wavenumber Domain Multiband Data Fusion'. Together they form a unique fingerprint.

Cite this