摘要： The polarimetric multiple-input multiple-output (MIMO) radar utilises both the polarisation and spatial diversities, thus the generalised likelihood ratio test (GLRT) of the polarimetric MIMO radar exhibits better detection performance than the GLRT of the MIMO radar when the secondary data set is large enough. However, the secondary data set is often not enough in reality. In this instance, the good detection performance of the GLRT of polarimetric MIMO radar cannot be guaranteed any more. In this study, the authors propose three kinds of generalised polarimetric whitening filters (GPWFs) to improve the detection performance of polarimetric MIMO radar in the limited secondary data environment. They first present the signal models and the test statistics for the three GPWFs. Then they analyse the detection performance of the three GPWFs by simulation, and find the second GPWF performs the best. They further compare the detection performance of the second GPWF with those of the GLRT and non-coherent integration (NCI) detector. Simulation results show that the proposed second GPWF outperforms the GLRT and NCI in the limited secondary data environment.