Distributed MIMO radar signal processing

Abstract

Radar systems are remote sensing tools that generate electromagnetic waves and extract information by receiving altered versions of these waves. Nowadays, many radar types are being used in specific areas such as weather prediction, automobiles, and the military. One type of radar employed in military applica-tions is called the multistatic radar system. Multistatic radar systems consist of multiple transmitters and receivers widely separated from each other. Although multistatic radar systems have not been invented recently, one type of multistatic radar has recently taken the attention of the literature, the multiple input mul-tiple output (MIMO) radar system. In this thesis, we analyze the performance of some techniques presented in the MIMO radar literature, make improvements, and propose new methods. First, we review the literature for MIMO radar waveform generation. Then, we propose a parameter estimation technique for multiple target cases using the polyphased-piecewise linear frequency modulated (PPLFM) waveform. Secondly, we propose a detection algorithm in which each receiver preprocesses received sig-nals and extracts bistatic range and Doppler for each transmitter. A grid of points in the region of interest (ROI) is generated, and by using a weighting function, a weight for each plot is calculated, and detection is performed via thresholding. Third, we propose a policy-iteration-based position and velocity estimation algo-rithm. We define a cost function using bistatic range and Doppler measurements for the proposed estimation algorithm. To perform estimation in the presence of multiple targets, we conduct data association by weighting the bistatic measure-ments. Fourth, a tracking algorithm that uses the Generalized Multi Bernoulli Filter is proposed. Lastly, we investigate the alternative MIMO antenna struc-tures and analyze the detection and tracking performance of the Electromagnetic Vector Sensor (EMVS). At the end of the thesis, it is demonstrated that the performance of the proposed algorithms is promising. Additionally, we show that the detection and tracking performance of the EMVS-based MIMO radar system is better than the performance of the MIMO radar system with dipole antennas.