A swap-based sparse array design method in personal sound zone

Area of Science:

• Acoustics and Signal Processing
• Optimization Algorithms
• Array Signal Processing

Background:

• Sparse array design (SAD) is crucial for enhancing personal sound zone (PSZ) performance.
• Existing methods may not adequately address multi-objective optimization in SAD for PSZ.
• Optimizing loudspeaker geometry is key to achieving desired acoustic fields.

Purpose of the Study:

• To formulate the SAD problem in PSZ as a sparse multi-objective optimization problem.
• To develop an efficient algorithm for solving this non-convex and combinatorial problem.
• To propose an approximate form for computational efficiency with multiple target sound fields.

Main Methods:

• Formulation of SAD in PSZ as a sparse multi-objective optimization problem considering multiple performance metrics.
• Design and proposal of a swap-based group search algorithm for near-optimal solutions.
• Development of an approximate form to reduce computational load for multiple target sound fields.

Main Results:

• The proposed method effectively controls trade-offs in PSZ applications.
• Achieved comparable or superior performance compared to existing approaches.
• The approximate form demonstrated effectiveness in reducing computational burden.

Conclusions:

• The novel multi-objective optimization approach and swap-based group search algorithm are well-suited for SAD in PSZ.
• The proposed method offers improved control over performance trade-offs.
• The approximate form provides an efficient alternative for specific scenarios.