Matched Illumination Waveforms using Multi-Tone Sinusoidal Frequency Modulation

TL;DR

This paper introduces a novel constant modulus waveform design using Multi-Tone Sinusoidal Frequency Modulation to optimize target detection in varying noise and clutter conditions, closely approximating ideal matched-illumination performance.

Contribution

The paper presents a new MTSFM-based method for designing constant modulus MI waveforms that better approximate ideal spectra compared to existing approaches.

Findings

MTSFM waveforms closely match ideal MI spectra in simulations.

MTSFM outperforms flat spectrum waveforms in detection performance.

Detection performance remains robust across varying noise and clutter PSDs.

Abstract

This paper explores the design of constant modulus Matched-Illumination (MI) waveforms using the Multi-Tone Sinusoidal Frequency Modulation (MTSFM) waveform model. MI waveforms are optimized for detecting targets in known noise and clutter Power Spectral Densities (PSDs). There exist well-defined information theoretic methods that describe the design of MI waveforms for a myriad of target/noise/clutter models. However, these methods generally only produce the magnitude square of the MI waveform's spectrum. Additionally, the waveform's time-series is not guaranteed to be constant modulus. The MTSFM is a constant modulus waveform model with a discrete set of design coefficients. The coefficients are adjusted to synthesize constant modulus waveforms that approximate the ideal MI waveform's spectrum. Simulations demonstrate that the MTSFM's detection performance closely approximates an ideal MI waveform spectrum and generally outperforms flat spectrum waveforms across a range of transmit energies when the noise and clutter PSDs vary greatly across the operational band.