Adaptive Transmit Waveform Design Using Multitone Sinusoidal Frequency Modulation

Abstract

This article presents an adaptive waveform design method using multitone sinusoidal frequency modulation (MTSFM). The MTSFM waveform's modulation function is represented as a finite Fourier series expansion. The Fourier coefficients are utilized as a discrete set of design parameters. These design parameters can be modified to shape the spectrum, auto-correlation function (ACF), and ambiguity function (AF) shapes of the waveform. The MTSFM waveform model naturally possesses the constant envelope and spectral compactness properties that make it well suited for transmission on practical radar/sonar transmitters which utilize high power amplifiers. The MTSFM has an exact mathematical definition for its time-series using generalized Bessel functions which allow for deriving closed-form analytical expressions for its spectrum, AF, and ACF. These expressions allow for establishing well-defined optimization problems that finely tune the MTSFM's properties. This adaptive waveform design model is demonstrated by optimizing MTSFM waveforms that initially possess a 'thumbtack-like' AF shape. The resulting optimized designs possess substantially improved sidelobe levels over specified regions in the range-Doppler plane without increasing their time-bandwidth product. Simulations additionally demonstrate that the optimized thumbtacklike MTSFM waveforms are competitive with thumbtacklike phase-coded waveforms derived from design algorithms available in the published literature.