The Journal of the Acoustical Society of America · 2017 · 59 citations · 20 references
Natural SciencesComputational EngineeringAcousticsComputational AcousticsOptimizationAcoustic ImagingAcoustic Inverse ProblemsAcoustic Signal ProcessingArray ProcessingBeamformingDigital Signal ProcessingAcoustic BeamformingAcoustic Source ReconstructionCompressive SensingMicrophone ArraysPerformance DegradationSemidefinite ProgrammingSound Field Reconstruction
Compressive beamforming realizes the direction-of-arrival (DOA) estimation and strength quantification of acoustic sources by solving an underdetermined system of equations relating microphone pressures to a source distribution via compressive sensing. The conventional method assumes DOAs of sources to lie on a grid. Its performance degrades due to basis mismatch when the assumption is not satisfied. To overcome this limitation for the measurement with plane microphone arrays, a two-dimensional grid-free compressive beamforming is developed. First, a continuum based atomic norm minimization is defined to denoise the measured pressure and thus obtain the pressure from sources. Next, a positive semidefinite programming is formulated to approximate the atomic norm minimization. Subsequently, a reasonably fast algorithm based on alternating direction method of multipliers is presented to solve the positive semidefinite programming. Finally, the matrix enhancement and matrix pencil method is introduced to process the obtained pressure and reconstruct the source distribution. Both simulations and experiments demonstrate that under certain conditions, the grid-free compressive beamforming can provide high-resolution and low-contamination imaging, allowing accurate and fast estimation of two-dimensional DOAs and quantification of source strengths, even with non-uniform arrays and noisy measurements.
20
Atomic Decomposition by Basis Pursuit
Scott Shaobing Chen, David L. Donoho, Michael A. Saunders · SIAM Review · 2001 · 5.1K citations
Compressed Sensing Off the Grid
Gongguo Tang, Badri Narayan Bhaskar, Parikshit Shah et al. · IEEE Transactions on Information Theory · 2013 · 1.2K citations
Natural Sciences, Electrical Engineering, Information Theory +16
The Convex Geometry of Linear Inverse Problems
2010 · 1K citations