FDA-MIMO Radar Target Detection with Limited Test Samples

Bang Huang,Ping Li,Jiangwei Jian,Wen-Qin Wang,Abdul Basit
DOI: https://doi.org/10.1109/taes.2024.3378201
IF: 3.491
2024-01-01
IEEE Transactions on Aerospace and Electronic Systems
Abstract:Since the dwell time of a frequency diverse array multiple-input multiple-output (FDA-MIMO) radar is very short and allows to generate a smaller set of useful transmitted samples than the data dimension under test (i.e., the positive covariance matrix can't be obtained), the overall detection performance deteriorates drastically. To address this challenge, we propose a reduced-dimensional (RD) approach in this paper. Specifically, based on the RD transformed data model, we introduce a detector utilizing the generalized likelihood ratio test (GLRT) criterion, named RD-GLRT. Furthermore, assuming that the transmit-receive vector (or Doppler vector) belongs to a given transmit-receive (or Doppler) subspace, we pre-multiply the test data with an appropriate matrix representing the transmit-receive subspace, resulting in a reduction in test data volume. Subsequently, we introduce two additional adaptive detectors, namely the RD subspace-based GLRT and the Wald criteria detectors, aiming for enhanced detection performance. Notably, all three proposed detectors exhibit a constant false alarm rate (CFAR) property against the covariance matrix and demonstrate a reduced demand for test data volume. Numerical results validate the effectiveness of these proposed detectors.
What problem does this paper attempt to address?