Diversity order and detection performance of MIMO radar: A relative entropy based study

Based on Steinpsilas lemma, we use relative entropy as an error exponent measure to study the detection performance of Neyman Pearson (NP) detectors. Both multiple-input-multiple-output (MIMO) radar and phased-array radar are investigated. We introduce the notion of diversity order to get the insights of results. For a bistatic radar where the transmitters and receivers are deployed separately with a fixed total number, the relative entropy measure also enables us to find the optimal fraction of receivers and transmitters. We find that for phased-array radar, the number of transmitters should always be equal to the number of receivers, while for MIMO radar, different configuration schemes should be adopted in different signal noise ratio (SNR) regions.