# An Iterative Fast Microphone Array Design Method Employing Equilateral Triangular Subarrays

## TL;DR

This paper introduces a new method for designing microphone arrays that is faster and more cost-effective by using triangular subarrays and iterative optimization.

## Contribution

The novel approach uses equilateral triangular subarrays and iterative genetic algorithm optimization to reduce design time and cost.

## Key findings

- The proposed method reduces optimization variables and computational domain significantly.
- Arrays designed with this method show improved main lobe width and sidelobe level performance near target frequencies.
- The method provides direct control over the number of array elements while optimizing frequency-specific performance.

## Abstract

In industrial acoustic imaging, microphone array design is often limited by the strong frequency dependence of array performance, the high computational cost of optimization, and the expense of deploying large numbers of microphones. Most existing optimization-based methods require simultaneous optimization of all array elements, resulting in long design times and limited flexibility in controlling element count. To overcome these limitations, this paper proposes a fast and iterative microphone array design method using equilateral triangular subarrays as basic units. Instead of optimizing the entire array at once, the proposed method incrementally adds subarrays, and in each iteration, a genetic algorithm optimizes only the placement of the newly added subarray for a specified target frequency. By exploiting the rotational symmetry of the equilateral triangular subarrays and the geometric characteristics of the array point spread function, the number of optimization variables and the computational domain are significantly reduced, enabling efficient array design. The proposed method allows frequency-specific performance optimization while providing direct control over the number of array elements, achieving a practical balance between imaging performance and hardware cost. Comparative results show that arrays designed using this method generally exhibit improved main lobe width and sidelobe level performance near the target frequencies compared with several classical array configurations.