Abstract:

Spatially colored noise would lead to performance degradation or even invalidation of multiple input multiple output (MIMO) radar algorithms. For the problem of joint direction of departure (DOD) and direction of arrival (DOA) estimation in bistatic MIMO radar with spatially colored noise, the reason for the failure of the existing algorithms is analyzed. Taking the low-rank property of the noiseless covariance matrix, the sparse feature of the colored noise covariance matrix as well as the multidimensional structure characteristic of the MIMO radar data after matched filtering into consideration, a tensor analysis-based angle estimation is introduced. Firstly, a covariance tensor for the angle estimation is constructed. The colored noise is suppressed via removing the entities that affected by the noise covariance measurement. Then the noiseless covariance tensor is recovered via tensor completion. Thereafter, the factor matrices corresponding to DOD and DOA are achieved via parallel factor (PARAFAC) decomposition. Finally, the DOD and DOA are fitted using least squares algorithm. Simulation results show that the proposed algorithm is not sensitive to the spatially colored noise, and it is free-from the aperture loss. The proposed algorithm performs more accurate estimation performance than the existing matrix and tensor approaches.

Key words: multiple input multiple output (MIMO) radar, direction of departure (DOD) and direction of arrival (DOA) estimation, spatially colored noise, tensor completion, parallel factor (PARAFAC) decomposition