Optimal Signal Design for a New Physically Motivated Clutter Model With Applications to Ultrasonic Testing.

Area of Science:

• Signal processing
• Ultrasonic imaging
• Waveform design

Background:

• Robust detectors are crucial for analyzing real-world ultrasonic data.
• Existing signal design methods may not fully leverage physically motivated clutter models.
• Analytical solutions for optimal signal design are often limited by data record assumptions.

Purpose of the Study:

• To explore the optimal signal design problem using a novel, physically motivated clutter model.
• To derive an analytical solution for the optimal signal and gain new insights into signal design.
• To investigate practical, realizable waveforms as alternatives to the theoretically optimal impulse signal.

Main Methods:

• Development of a new physically motivated clutter model for ultrasonic applications.
• Derivation of an analytical solution for the optimal signal, proving its validity for finite data records.
• Comparative analysis of various waveforms (LFM, NLFM, PCM, GAMS) for detectability and performance under complex target scenarios.

Main Results:

• The optimal signal for the proposed model is theoretically an impulse with an impulsive autocorrelation sequence (ACS).
• Linear frequency-modulated (LFM) pulses demonstrate a significant advantage in detectability over the Gaussian amplitude-modulated sinusoid (GAMS) pulse.
• LFM and GAMS signals exhibit different performance characteristics when deviating from the single scatterer assumption with simulated noise.

Conclusions:

• The proposed clutter model facilitates robust detector design and provides analytical solutions for optimal signal waveforms.
• While an impulse is theoretically optimal, LFM pulses offer a practical and superior alternative to GAMS pulses for ultrasonic applications.
• Further research into waveform performance under complex target conditions is warranted for enhanced ultrasonic system design.