Gain Assessment of Environmental-Informed Subarray Processing: The ALMA 2017 Example
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Underwater acoustic propagation can be affected by ocean fluctuations. In particular, internal waves can reduce the spatial coherence of the acoustic signal on SONAR arrays, causing performance degradation. Subarray processing is a popular method, used to mitigate the performance loss caused by environmental fluctuations. The problem lies in the correct parameterization of the length of the subarrays, itself linked to the acoustic coherence length, with the latter influenced by the fluctuations of the environment. In this work, we propose the use of a linear model, derived from canonical correlation analysis (CCA), to configure the length of the subarrays according to statistics of acoustic and environmental measurements, in the context of direction-of-arrival estimation. To compensate for the loss of angular resolution inherent to the combination of subarrays using conventional beamforming, we propose to consider a sparse regularization, solved using SoBaP, a Bayesian algorithm. We show on at-sea data from the 2017 ALMA campaign that regularization compensates for some of the resolution loss due to the use of the subarrays. In addition, detection performance is increased when using the regularized subarray scheme, particularly when acoustic coherence is low. Finally, using the CCA-informed scheme shows better detection performance than the noninformed schemes for low coherence and low signal-to-noise ratio, demonstrating the benefits of informing array-processing algorithms with environmental statistics.
Fields of Study & Subjects
Direction-of-Arrival Estimation Techniques
Underwater Acoustics Research
Underwater Vehicles and Communication Systems
Publication details
- Journal / source
- IEEE Journal of Oceanic Engineering
- Volume
- 51
- Issue
- 2
- Pages
- 916–927
- Publication date
- 30 March 2026
- Language
- EN