Metaheuristic algorithms in analog integrated circuit design

Development in the technology increases demand to high efficiency and more precise analog integrated circuits. Nowadays analog integrated circuit design is generally managed by optimization tools. Thus, usage of optimization tools in order to have high precision from analog integrated circuits has gained more importance than they were in the past. Developing reliable automatic tools in analog IC design is very important because of the optimization contradictions. This thesis presents the benefits of evolutionary and multi-objective algorithms which are metaheuristic algorithms on a two stage operational amplifier design using Harmony Search Algorithm (HSA) and Non-dominated Sorting Genetic Algorithm (NSGA-II). HSA is a new kind of algorithm which was inspired from the musicians those are looking for a perfect state of harmony that produces the best sound. NSGA-II is one of the most efficient algorithms using elitist approach which sorts the population in different fronts using the non-domination order relation. Synthesis of the opamp is realized with these algorithms. There are several contradictions in the opamp design. For instance, in order to decrease the power dissipation, the tail current should be decreased. However reducing the current causes a decrease in slew rate and gain bandwidth product. Thus, optimization is required to find an optimum point. This optimization problem was solved using Harmony Search and Non-dominated Sorting Genetic Algorithms in order to obtain the optimum transistor sizes and optimum values of other circuit parameters which satisfy all the constraints.