Target Scattering Diversity Property And Detection Algorithms For Diversity MIMO Radar

Multiple-input multiple-output (MIMO) becomes a hot topic in the radar field recently. It can use the diversity of transmitted signals to improve radar performances. According to the distance between radar antennas, MIMO radar is categorized into two types, MIMO radar with collocated antennas and that with widely separated antennas. The former can use the signal waveform diversity to improve the degrees of freedom and angular resolution of a radar system, while the later can use the target spatial scattering diversity to improve the target detection performance, hence called spatial diversity radar. Another type of diversity radar is frequency diversity radar. The spatial-frequency jointly diversity MIMO radar, which uses both the spatial diversity technique and the frequency diversity technique, gains increasing attention. Supported by the advanced defence research program of China, named“XXX/XXX radar technique”, this work investigates signals processing and target detection algorithms for collocated MIMO radar and distributed MIMO radar. There are mainly four contributions in the work.The statistical characteristic of radar target echoed signals, especially the target correlation characteristic, is an important concern in designing target detection algorithms for a diversity MIMO radar. It will be shown that a diversity MIMO radar may receive target echoed signals that are neither completely correlated nor independent. The widely used independence criterion becomes incapable of describing the exact degree of correlation of target echoed signals anymore. Therefore, at the background of a spatial-frequency jointly diversity MIMO radar, the correlation of target echoed signals is studied. An expression of a correlation coefficient expression to describle the correlation of target echoed signals in arbitrary two diversity channels therein is developed here. This result is applicable to any spatial diversity radar, frequency diversity radar and spatial-frequency jointly diversity MIMO radar. The expression can be used to estimate the correlation coefficient between echoed signals of a possible target, which can be used to design novel detection algorithms and evaluate existing detection algorithms.A MIMO radar with collocated antennas can reap more degrees of freedom and achieve a higer angular resolution, at the cost of more computational cost. Based on a fact that the interference covariance matrix (ICM) may take a Kronecker production structure, an adaptive target parameter estimation algorithm and two typical signal processing schemes are proposed, which can reduce the samples size requirement and computational cost. Subsequently, the ICM structures of typical interferences are analyzed, indicating that inactive interference would spoil this structure. Via the MIMO beamforming problem, the impact of inactive interferences on MIMO beamforming algorithms based on the ICM structure is studied, indicating that such an algorithm can suppress inactive interferences fewer than receiving antennas, but can not suppress more inactive interferences.Diversity MIMO radar can improve target detection performance by fusing signals received in multiple diversity channels. The mutual corelaiton and signal noise ratio (SNR) distribution are two major concerns in designing signal fusion based target detection algorithms for diversity MIMO radar. Their impacts on the detection performances of existing detection algorithms are studied. Moreover, based on the Newman-Pearson criterion, an SNR-weighting based detection algorithm is proposed to detect signals of different channel SNRs, and a general Gaussian signal detection algorithm to detect signals containing target echoed signals statistically correlated and of different SNRs. The impacts of the target correlation, channel SNR-unbalance and false alarm rate on concerned target detection algorithms are studied, indicating that the target correlaiton and channel SNR-unbalance can degrade conventional detection algorithms, while the proposed algorithms can improve the radar detection performance.A practical diversity MIMO radar detection system should suppress clutter signals received in local radar sites. Based on the generalized likelihood ratio test (GLRT) criterion, a signal fusion based target detection problem with the local clutter suppression capability is proposed. The detector has many properties. It can detect targets at a constant flase alarm rate. It can be used to conveniently organize a signal fusion based radar network that can allow local covariance matrices of clutter returns receiced in widely separated radar sites to be different and local radar parameters to be flexibly set. Relay nodes therein can use passby data to improve its detection performance, without increase of the communication bandwidth when data go through relay nodes.