FAst optimization of analog amplifier architecture using simulated annealing

In case of architectural designing of an analog system/sub-system, accurate and sufficient estimation of non-idealistic parameters are of highest importance as it impacts strongly on the cost and performance of the target implementation. Manual optimization of such a system which is a multi-dimensional complex design-space, is prone to manual errors, limited by human imaginations and should be avoided. Global-optimization algorithms like Genetic Algorithm (GA) or Simulated Annealing (SA) fails to find out the required minima in such a complex design space. In this article we present a novel method to optimize Programmable Gain Amplifier (PGA) as an analog system and/or sub-system accurately by addressing the limitations of SA and by using refined SA as one of the chosen global optimization algorithm. Proposed method is fast, reliable and easy to use for the architect community.