Application and Resulting Suitability of a Genetic Algorithm in the Design of FGMOS-based CMOS-MEMS Transducers

In this work an initial approach and requirements overview is performed in order to promote awareness on the compatibility between CMOS-MEMS sensor devices design and the group of heuristic techniques known as genetic algorithms. As might be known, genetic algorithms (GAs) main application is in the field of multivariable functions optimization and this kind of iterative procedures in their simplest forms may be suitable to serve as a tool in the automation of design of integrated CMOS devices and technologies where the variables and restrictions are known. FGMOS-based devices along with MEMS structures when embedded in a single-chip platform (CMOS-MEMS technology) are expected to be in compliance with a set of design rules for both their mechanical and electrical properties, making easier to code and decode variables for their use in a GA in a discrete always-positive basis. In this approach, parasitic and process-resolution-related issues are neglected and further analysis based in a more detailed modeling and parameter restrictions is encouraged.