Micro Doppler Reconstruction From Discontinuous Observations Based on Gapped SBL-FBTVAR Method for Spin Stabilized Object

Abstract

Micro Doppler analysis of spin stabilized objects is of a great significance for attitude estimation and recognition of space targets. In practice, the radar cannot dwell on one target in a long interval continuously. In this paper, we propose a novel approach for the micro Doppler frequency recovery from the discontinuous radar observations, which is referred to as gapped sparse Bayesian learning forward backward time-varying autoregressive (GSBL-FBTVAR) method. First, the sparse optimization model for estimating the sparse FBTVAR model coefficients from gapped samples is established. Then, the sparse FBTVAR model parameters corresponding to the gapped sampled data and the missing data are estimated via an extended sparse Bayesian learning (SBL) algorithm and the missing-data iterative adaptive approach (MIAA). The micro Doppler frequencies are estimated by investigating the relationship between the model parameters and the poles. Finally, the experiments are carried out on the electromagnetic analysis data to verify the proposed GSBL-FBTVAR method.

Keywords

• Spin stabilized object
• micro Doppler recovery
• gapped samples
• TVAR
• sparse Bayesian learning