Steerable sparse linear array design based on compressive sensing with multiple measurement vectors.

Area of Science:

• Acoustics
• Signal Processing
• Array Signal Processing

Background:

• Underwater acoustic imaging systems often use non-uniform arrays to achieve desired performance.
• Compressive Sensing (CS) methods offer a way to synthesize sparse arrays, reducing element count while maintaining imaging capabilities.
• Traditional single measurement vector (SMV) CS methods can introduce artifacts in beam patterns when arrays are steered off-broadside.

Purpose of the Study:

• To synthesize a non-uniform linear array for two-dimensional Range-Bearing underwater acoustic imaging.
• To address the limitations of single measurement vector (SMV) CS in sparse array synthesis for sonar systems.
• To demonstrate the efficacy of the multiple measurement vectors (MMV) CS method for improved beam pattern performance.

Main Methods:

• Utilized the multiple measurement vectors (MMV) compressive sensing (CS) method for sparse linear array (SLA) synthesis.
• Conducted simulation studies for Forward Looking Sonar systems with multiple beams in the field of view (FOV).
• Compared the MMV CS synthesized SLA with an SLA synthesized using single measurement vector (SMV) CS.

Main Results:

• The MMV CS method effectively synthesizes sparse linear arrays (SLAs) for underwater acoustic imaging.
• SMV CS synthesized SLAs exhibited spurious lobes when steered away from broadside, degrading imaging performance.
• MMV CS synthesized SLAs demonstrated superior beam pattern characteristics, overcoming the limitations of SMV CS.

Conclusions:

• The MMV CS method is a robust approach for designing sparse linear arrays (SLAs) for underwater acoustic imaging.
• This technique enhances the performance of Forward Looking Sonar systems by mitigating beam pattern artifacts.
• The study provides designs for SLAs across various frequencies and angular resolutions, applicable to diverse sonar applications.