A low-complexity robust estimation of multiple wideband polynomial-phase signals in sensor array

Abstract

This paper addresses joint estimation of parameters of multiple polynomial-phase signals (PPSs) impinging on a sensor array and corresponding direction-of-arrivals (DOAs). Using a recently proposed method for the fine estimation of multiple PPSs, we first separate signal components and then estimate the parameters of interest from the spatial high-order instantaneous moments (SHIMs) of the obtained components. The parameter estimation is performed through maximizing the spectrum obtained by averaging the spectra of multiple rows and columns of the SHIM matrix. The proposed approach offers lower signal-to-noise ratio (SNR) threshold and higher accuracy with respect to concurrent methods. Since all parameters are estimated through 1-D searches, the calculation complexity is significantly reduced with respect to the maximum likelihood (ML) methods.