Reduced-rank sub-array detection and the optimal sub-array division for spatial correlation attenuation

In this paper, we consider the problem of detection performance degradation caused by the spatial correlation attenuation in the ocean environment. A reduced-rank detector is developed via combining the subspace Eigen Value Decomposition(EVD) with the sub-array processing, and the performance of the detector is evaluated. The results show that the reduced-rank detector using sub-arrays has a better performance than the full-array detector in the presence of imperfect correlation. Meanwhile, effects on the sub-array detection performance of the sub-array geometry are studied, and the optimal sub-array division method is proposed. We notice that there is a certain proportional relation between the optimal sub-array length and the signal correlation length, and that the optimal detection performance can be reached, as the ratio between the sub-array length and the correlation length is in the range of 1 to 2.5. The results are validated by computer simulations.