Time-varying interference suppression in communication systems using time-frequency signal transforms

Abstract

The Fourier transform (FT) is an analysis tool that is perfectly localized at sinusoidal frequencies. We present a FT generalization (GFT) as a 1-D transform of the frequency-modulation (FM) rate. It is designed to be perfectly localized at the FM rate of signals with linear or non-linear instantaneous frequency in the time-frequency plane. We discuss various important GFT properties, implementation and aliasing issues, and provide examples. We successfully apply the GFT to suppress broadband linear FM interference in spread spectrum communication systems. We demonstrate that this new method also performs well for multicomponent interference, and has a computational advantage over other time-varying techniques. In addition, we show that some examples of the GFT are adequate in reducing dispersive interference in underwater and satellite communication systems.