Matched insonified waveform design using multi-tone sinusoidal frequency modulation

Matched Insonified (MI) active sonar transmit waveforms leverage information theoretic methods to optimize their Energy Spectral Densities (ESD) for optimal object detection in environments with estimatable noise and clutter Power Spectral Densities (PSD). While these methods determine the optimal waveform’s ESD, they do not specify a time series that realizes that optimal ESD shape. The Multi-Tone Sinusoidal Frequency Modulated (MTSFM) waveform is an adaptive waveform model that controls its spectral shape with a discrete set of coefficients. Manipulating these coefficients allows MTSFM waveforms to approximate optimal MI waveform ESDs while maintaining a constant modulus, a critical waveform property that facilitates efficient transmission on high-power amplifiers. This research evaluates the detection performance of MTSFM waveforms for several hollow elastic objects in a physics based acoustic simulation with variable noise and clutter PSDs. Receiver Operating Characteristic (ROC) curves compare the detection performance of ideal MI waveforms, MTSFM waveforms approximating the MI waveform’s ESD, and flat spectrum waveforms with transmit energy and bandwidth constraints. The MTSFM based MI waveforms closely approximate the performance of ideal MI waveforms and generally outperform flat spectrum waveforms when the object, noise, and clutter spectral content varies greatly across the frequency band of interest.