Improving the Efficiency of DAMAS for Sound Source Localization via Wavelet Compression Computational Grid (1608.05179v2)

Abstract: Phased microphone arrays are used widely in the applications for acoustic source localization. Deconvolution approaches such as DAMAS successfully overcome the spatial resolution limit of the conventional delay-and-sum (DAS) beamforming method. However deconvolution approaches require high computational effort compared to conventional DAS beamforming method. This paper presents a novel method that serves to improve the efficiency of DAMAS via wavelet compression computational grid rather than via optimizing DAMAS algorithm. In this method, the efficiency of DAMAS increases with compression ratio. This method can thus save lots of run time in industrial applications for sound source localization, particularly when sound sources are just located in a small extent compared with scanning plane and a band of angular frequency needs to be calculated. In addition, this method largely retains the spatial resolution of DAMAS on original computational grid, although with a minor deficiency that the occurrence probability of aliasing increasing slightly for complicated sound source.

• The paper demonstrates that wavelet compression grids reduce DAMAS computation time by up to 99.6% for single-point source configurations.
• It introduces a data sparsity approach that retains only crucial grid points, significantly trimming redundant computations.
• The method maintains spatial resolution across various source configurations, making it viable for real-time acoustic localization.

Enhancing DAMAS Efficiency for Acoustic Source Localization Through Wavelet Compression

The paper authored by Wei Ma and Xun Liu explores the use of wavelet compression to improve the efficiency of the DAMAS (Deconvolution Approach for the Mapping of Acoustic Sources) algorithm, particularly in applications involving sound source localization. Phased microphone arrays are crucial in such areas, and while DAS (Delay-and-Sum) beamforming offers a robust solution, its spatial resolution is typically limited. DAMAS, introduced by Brooks and Humphreys, overcomes this by deconvolving the beamforming map iteratively to achieve improved spatial resolution. However, the computational demands of DAMAS can be prohibitive, which spurred the authors to explore methods for enhancing its computational efficiency without altering the core DAMAS algorithm.

Key Contributions

The authors propose using a wavelet compression computational grid to substantially reduce the computational effort required by the DAMAS algorithm, especially when sound sources are clumped in a small region compared to the larger scanning plane. The wavelet compression identifies and retains only essential grid points, which represent the significant information about sound sources, thus discarding redundancies. This strategy leverages the natural data sparsity in the beamforming map to construct a more efficient computational grid.

Numerical Results and Implementation

Several application simulations are discussed in the paper, utilizing a computational setup with a planar microphone array. The simulations include variations of point source configurations—single and multiple points with equal or differing acoustic power levels. Key outcomes of the simulations illuminate the efficiency gains achievable with wavelet compression grids:

• Efficiency Improvement: The application of wavelet compression grids led to significant reductions in DAMAS computation time. For instance, a single grid point source configuration saw a 99.6% efficiency gain. The authors measure run time reductions quantitatively, demonstrating that compression significantly alleviates computational demands.
• Spatial Resolution Retention: Across varying configurations, including two-point sources with different sound powers and more complex arrangements (such as DAMAS image sources), the method retained the spatial resolution offered by traditional DAMAS. The preservation of spatial resolution with an efficiency gain underlines the practical viability of the wavelet compression approach.

Implications and Future Directions

This work substantiates the utility of integrating wavelet compression grids with the DAMAS approach, extending its practical utility in real-time applications where computational efficiency and spatial resolution are paramount. The findings suggest relevant practical implications in industrial environments where precise acoustic localization is needed over large frequency bands.

The paper also touches on potential areas for future exploration, such as optimizing the computational grid further for other deconvolution algorithms like DAMAS2 or FFT-NNLS, which might benefit similarly from this wavelet-based compression strategy. Moreover, extending the approach to 3D acoustic imaging represents an opportunity for substantial impact, particularly in aerospace engineering where spatial acoustic measurements are becoming more sophisticated.

In summary, the authors present a compelling case for transforming the utility of DAMAS through computational grid optimization. This exploration into wavelet compression not only positions DAMAS as a more efficient tool for acoustic source localization but also provokes further discourse on adaptive grid methods that might enable new capabilities in computational acoustics.