Signal processing techniques for target detection and estimation for radar systems

Abstract

With the continuous increase in cases of enemy attacks, road accidents, goods traf ficking, and burglary attempts, surveillance is necessary for the safety of the country, human wealth, and property. Since the inception of RADAR (during the second World War), which is an acronym for radio detection and ranging, radar has been extensively used for surveillance and monitoring. In-fact with continuous technological advancement and because of the advent of various digital signal processing schemes, researchers are continuously working to enhance the performance of radar systems. The two essential aspects of radar systems are detection and estimation. The detector gives an estimate of the number of targets present in the surveillance environment. Subsequently, the estimator yields the estimates of the location and speed of the detected targets. Practically, the reflected signal processed for the detection of targets and estimation of their range and velocity, are heavily perturbed by thermal noise and clutter. Consequently, perturbation because of thermal noise and clutter hinders the perfect detection of targets and accurate estimation of their location and velocity. The goal of this thesis is to propose new techniques for target detection and estimation of the target’s range and velocity by suppressing the effects of thermal noise and clutter. For this purpose, in this thesis, the following works are done. Firstly, the orthogonal frequency division multiplexing (OFDM) based radar system is explored for the detection of a small boat in sea clutter. For this, we propose a technique to generate radar return for OFDM waveform using collected radar return data for stepped frequency waveform. We then derive the system model for the estimated radar return data specific to the OFDM waveform. Further, a detection test is proposed for the derived signal model and surveillance environment. The close match between the derived analytical expressions and simulation results validates the performance of the proposed detector. For estimation of the target’s range and velocity, an adaptive estimator based on a sparse kernel least mean square algorithm is proposed. Being an adaptive algorithm, the estimates are obtained by low computational complexity, and the accuracy of estimates is guaranteed by the convex nature of optimizing cost function in reproducing kernel Hilbert space. Subsequently, an adaptive kernel width optimization technique is proposed to further lower the computational complexity of the proposed estimator. An expression for the Cramer-Rao lower bound (CRLB) is derived and validated for the proposed estimator over linear frequency modulated, and OFDM radar systems. In the next work, we propose kernel maximum correntropy based estimators for range and velocity estimation in non-Gaussian clutter. Additionally, an adaptive update equation is derived for optimization of the kernel-width, which further lowers the dictionary-size, and variance of the proposed estimator. For the performance evaluation of the proposed estimators, an expression is derived for the CRLB using a modified Fisher information matrix (FIM). Next, a kernel minimum error entropy (KMEE) based estimator is proposed for the es timation of multiple targets’ direction of departure (DOD), the direction of arrival (DOA), and the Doppler shift with multiple input multiple output radar in non-Gaussian clutter. The computational complexity of the proposed KMEE based estimator is reduced by in corporation of novelty criterion based sparsification technique. Analytical expressions are derived for the variance of estimation-error in DOD, DOA, and Doppler shift. Further, for assessing the accuracy of the proposed estimator, the CRLB is calculated using the Modified FIM. Lastly, two efficient low complexity estimators, namely, the extended Kalman filter (EKF) and the unscented Kalman filter (UKF), are proposed. The EKF is advantageous due to its implementation simplicity and fast computation; however, a derivative-based implementation limits its use. The UKF outperforms the EKF and offers better stability due to a derivative-free implementation. Simulation results reveal improved accuracy achieved by the proposed EKF and UKF based estimators. Moreover, the EKF and UKF based estimators show a closer match with the CRLBs compared with the existing approaches along-with low computational complexity.