The topic of multiple input multiple output (MIMO) radar recently gained considerable interest because it can transmit partially correlated or fully independent waveforms. The inherited waveform diversity helps MIMO radars identify more targets and adds flexibility to the beampattern design. The realized advantages come at the expense of enhanced processing requirements and increased system complexity. In this regards, a closed-form method is derived to generate practical finite-alphabet waveforms with specific correlation properties to match the desired beampattern. Next, the performance of adaptive estimation techniques is examined. Indeed, target localization or reflection coefficient estimation usually involves optimizing a given cost-function over a grid of points. The estimation performance is directly affected by the grid resolution. In this work, the cost function of Capon and amplitude and phase estimation (APES) adaptive beamformers are reformulated. The new cost functions can be evaluated using the two-dimensional fast-Fourier-transform (2D-FFT) which reduces the estimation runtime. Generalized expressions of the Cram´er-Rao lower bound are computed to assess the performance of our estimators. Afterward, a novel estimation algorithm based on the monopulse technique is proposed. In comparison with adaptive methods, monopulse requires less number of received pulses. Hence, it is widely used for fast target localization and tracking purposes. This work suggests an approach that localizes two point targets present in the hemisphere using one set of four antennas. To separate targets sharing the same elevation or azimuth angles, a second set of antennas is required. Two solutions are suggested to combine the outputs from the antenna sets and improve the overall detection performance. The last part of the dissertation focuses on the application and implementation side of radars rather than the theoretical aspects. It describes the realized hardware and software design of a compact portable 24 GHz frequency-modulated-continuous-wave (FMCW) radar. The prototype can assist the visually impaired during their outdoor journeys and prevents collisions with their surrounding environment. Moreover, the device performs diverse tasks such as range-direction mapping, velocity estimation, presence detection, and vital sign monitoring. The experimental result section demonstrates the device’s capabilities in different use-cases.
|Date of Award||Nov 2018|
- Computer, Electrical and Mathematical Science and Engineering
|Supervisor||Mohamed-Slim Alouini (Supervisor)|