Greedy Array Geometry Optimization Based on Maximin Principle for Region Sound Acquisition

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This paper presents a method to improve sound acquisition in microphone arrays by optimizing their geometry. We aim to have a constant array gain in the target region by maximizing the minimum full-band array gain, iteratively and sequentially. To find the optimal array geometry, we develop a novel greedy algorithm with two main steps: forward addition and backward deletion. In the forward addition step, we select a new sensor from a pool of candidates to enhance the array gain in the minimum direction identified in the previous step. In the backward deletion step, we remove a sensor that contributes the least after several forward additions. Simulation results indicate that the full-band array gain of the optimized array remains nearly constant within the desired region. Additionally, the beampattern of the optimal array exhibits good rotational invariance in this region.

Original languageEnglish
Title of host publicationProceedings of 2025 IEEE 15th International Conference on Signal Processing, Communications and Computing, ICSPCC 2025
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9798331565466
DOIs
StatePublished - 2025
Event15th IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2025 - Hong Kong, China
Duration: 18 Jul 202521 Jul 2025

Publication series

NameProceedings of 2025 IEEE 15th International Conference on Signal Processing, Communications and Computing, ICSPCC 2025

Conference

Conference15th IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2025
Country/TerritoryChina
CityHong Kong
Period18/07/2521/07/25

Keywords

  • Greedy algorithm
  • array geometry optimization
  • region sound acquisition

Fingerprint

Dive into the research topics of 'Greedy Array Geometry Optimization Based on Maximin Principle for Region Sound Acquisition'. Together they form a unique fingerprint.

Cite this