Inspired by the recent advantages of Multiple-Input Multiple-Output (MIMO) technologies in wireless communications, the MIMO concept was adopted in the radar context. By exploiting the potentials of MIMO techniques in respect of combating the fading of a channel, a MIMO radar system was created to handle a similar problem by exploiting widely separated antennas, namely, the variations in the returned signal power from different aspects of a target of interest. Compared with conventional radar systems, such a multiple-antenna radar system may result in a better understanding or estimation of a target’s radar cross section by angular or spatial diversity, which consequently delivers a better performance.
This dissertation studied the detection performance of MIMO radar systems: the study can be divided into two parts, viz. detection of stationary targets and detection of moving targets.
With respect to the detection of stationary targets, firstly, the detection performance of MIMO radar systems in white Gaussian noise was examined. A comparison between MIMO and phased array radar systems was then presented to illustrate their respective properties. Secondly, the robustness of MIMO radar systems to clutter was investigated. It was found that MIMO radar systems are better able to remove the effects of clutter than that of phased array radar systems. Thirdly, a more practical distributed MIMO radar system in respect of bandwidth limitations was proposed. Three classical distributed algorithms, namely, OR, AND, and Majority Logic (MAJ), were applied to existing MIMO radar systems. Lastly, a MIMO passive radar system based on FM waveform was introduced. By using the Neyman-Pearson hypothesis, an optimal receiver for the MIMO passive radar was developed.
Taking Doppler shift into account, detection performance of moving targets was investigated. The effects of instantaneous Doppler shift on the detection performance were first examined. Afterwards, the detection of moving targets including range considerations was discussed.
Additionally, the pertinent Matlab source codes used throughout the dissertation are attached. These will assist both newcomers and current radar researchers who study the detection performance of MIMO radar systems.