Target detection in distributed MIMO radar with registration errors

Abstract

In this paper, we consider a target detection problem in a distributed multiple-input multiple-output radar with registration errors. Registration errors mean the alignment errors when observation data from different distributed radars are transformed in a common coordinate system. Our study is motivated by the fact that the perfect registration condition is unavailable in practice, even after a registration process. A model for the imperfect registration is introduced where an uncertainty window centered on the cell being tested along range and azimuth in each spatial diversity channel is employed. A generalized likelihood ratio test (GLRT) detector and a Bayesian likelihood ratio test detector are developed based on the uncertainty windows. The probability of false alarm of the proposed GLRT detector is derived. For convenience of comparison, we also analyze the detection performance of the conventional likelihood ratio test (LRT) detector with registration errors. Simulation results demonstrate that the detection performance of the two proposed detectors outperform the conventional LRT detector in the imperfect registration case.