Supervisor Chalmers: Prabhat Khanal
Supervisors SAAB: Milena Anguelova & Bengt Svensson
Examiner: Rob Maaskant
The synthesis of isophoric sparse antenna arrays is explored using a genetic-inspired non-hard constrained optimization routine. To achieve the desired transmit beamforms, antenna element positions and phases are optimized. The elements are grouped into identical fixed array submodules, the placement of which in addition to their orientation is optimized. The modular design is proposed in order to reduce manufacturing difficulties and costs associated with N free roaming elements. The component sizes at high frequency motivate a minimal element spacing constraint. Expanding upon previous work, new computationally efficient coupling models are explored. Hinted by experimental studies, a radial-dependent mutual-coupling model is semi-empirically derived using data from a full-wave solver. While the locality of self-coupling is explored using a neural network and a regression model. The coupling models yields improved reproducibility with a full-wave solver compared with the previous study, at minimal computational expense. The modular design shows only a small performance impact compared to its free roaming counterpart.
Ludvig and Rob
E2 Room 7430 Landahlsrummet