| Title | Spatial-Resampling Wideband Compressive Beamforming |
| Publication Type | Conference Paper |
| Year of Publication | 2021 |
| Authors | Zhu, Zhen, Chi, Cheng, Zhang, Chunhua |
| Conference Name | OCEANS 2021: San Diego – Porto |
| Keywords | array signal processing, Broadband communication, composability, Compressive beamforming, compressive sampling, direction-of-arrival estimation, estimation of the direction of arrival, Interference, pubcrawl, resilience, Resiliency, Robustness, Sonar applications, Spatial resolution, Spatial-Resampling, wideband models |
| Abstract | Compressive beamforming has been successfully applied to the estimation of the direction of arrival (DOA) of array signals, and has higher angular resolution than traditional high-resolution beamforming methods. However, most of the existing compressive beamforming methods are based on narrow signal models. Wideband signal processing using these existing compressive beamforming methods is to divide the frequency band into several narrow-bands and add up the beamforming results of each narrow-band. However, for sonar application, signals usually consist of continuous spectrum and line spectrum, and the line spectrum is usually more than 10dB higher than the continuous spectrum. Due to the large difference of signal-to-noise ratio (SNR) of each narrow-band, different regularization parameters should be used, otherwise it is difficult to get an ideal result, which makes compressive beamforming highly complicated. In this paper, a compressive beamforming method based on spatial resampling for uniform linear arrays is proposed. The signals are converted into narrow-band signals by spatial resampling technique, and compressive beamforming is then performed to estimate the DOA of the sound source. Experimental results show the superiority of the proposed method, which avoids the problem of using different parameters in the existing compressive beamforming methods, and the resolution is comparable to the existing methods using different parameters for wideband models. The spatial-resampling compressive beamforming has a better robustness when the regularization parameter is fixed, and exhibits lower levels of background interference than the existing methods. |
| DOI | 10.23919/OCEANS44145.2021.9706024 |
| Citation Key | zhu_spatial-resampling_2021 |
Spatial-Resampling Wideband Compressive Beamforming