High-resolution sensor array processing in the presence of multiple wideband chirp signals

Estimating the parameters of polynomial-phase and chirp signals in sensor arrays is an important task which is frequently encountered in practical applications. Several authors have approached this problem using the narrow-band setting. In this paper, we present an optimal (maximum likelihood) algorithm for estimating the direction-of-arrival (DOA) and frequency parameters of multiple wideband constant-amplitude polynomial-phase signals. Since the proposed ML estimator is computationally intensive, an approximate solution is considered, originating from the analysis of the likelihood function in the single polynomial-phase signal case. As a result, the so-called polynomial-phase beamformer is obtained. Its simplified version referred to as the chirp beamformer is considered in detail. Explicit expressions for the corresponding Cramer-Rao bound (CRB) are presented as well. The performances of the exact ML algorithm and the chirp beamformer are compared to the CRB.