الفهرس 
INTRODUCTIONOnly 14 pages are availabe for public view
 |  | from | 146 |  |  |
| | | from | 146 | | |
Abstract
Automated parasitic-aware simulation-based optimization design tool is presented in this
thesis. Knowledge-aware simulation-based optimization and weighting factor methodologies are
proposed to reach an optimal and speedy design. The optimization core uses Broyden–Fletcher–
Goldfarb–Shanno (BFGS) optimization methodology. Parasitic estimation is done before the
sizing process to speed up the design and help obtain an optimum and efficient design. This
parasitic estimation is done via a parasitic model generator which models both the interconnects’
effects on active devices and the non-linearities of passive devices. Compared to the state-of-theart tools, the proposed tool gives limitless solutions when compared to template-based
algorithms and a guaranteed converged solution compared to random optimization ones. The tool
works in a wide range of RF frequencies. It uses the design of mm-wave front end receiver
circuits of phased array systems, such as low noise amplifier (LNA), N-bits digitally controlled
active phase shifter (PS), and current steering variable gain amplifier (VGA) using the 65nm
CMOS technology as a validation of its efficiency. To our knowledge, it is the first reported
automated tool used to design current steering VGA and active PSs. The output of the tool is
generated at the output interface. The outputs are as listed, the circuit schematic, the optimal
design parameters, and a generic layout with the GDSII file for further modifications.