Computational Efficiency of Circuit Design and Optimization Algorithms: A Comparative Study

Design of analog circuits is a computationally-complex process. Therefore, robust, simple, and efficient algorithms are required for design automation purpose. Developing generalized analog design methodologies, valid across a large solution space, is very demanding. This article compares computational complexity and precision of three different design methodologies, i.e., g(m)/I-D, Inversion Coefficient, IC, and C/I-D, or Inverse-ID approaches. The three methodologies will be analyzed to explore their performance to meet basic needs of the circuit design industry. A basic analog circuit block, i.e., a common-source amplifier, has been selected for this analysis. This study shows that the C/I-D method requires the least approximation for optimization.