Maximum likelihood and suboptimal schemes for micro-Doppler estimation using carrier diverse Doppler radars

Carrier diverse radars, known as dual-frequency radars, employ two different frequencies, and can be effective in determining the moving target range in urban sensing and through-the-wall radar applications. The authors derive the maximum likelihood (ML) estimator for the micro-Doppler motion parameters from the dual-frequency radar returns. Micro-Doppler signatures, which are commonly associated with vibrating, oscillating and rotating objects, have emerged to be an important tool in target detection and classification. Unlike linear models, the respective ML estimator does not assume a closed-form expression. The authors solve the ML estimator for dual-frequency radar operations by applying an iteratively reweighted non-linear least squares algorithm (IRNLS), which is initiated using suboptimal solutions. The ML-IRNLS algorithm is applied to both simulated and experimental radar returns for estimating the range and the motion parameters of indoor targets.