Design Automation, Modeling, Optimization, and Testing of Analog/RF Circuits and Systems by Particle Swarm Optimization

2.1 Automatic Circuit Design: In Loop Optimization with CAD

PSO, when clubbed with CAD tools, is used for automatic electronic design of Analog/RF circuits. In this process, an objective function is defined in the design tool and based on this objective function PSO is applied. PSO can be implemented in the design tool itself or in any other computational tool which can communicate with the design tool. Based on the objective function, the design is iterated by PSO algorithm. This can be called as in-loop optimization because for each particle CAD tool needs to be run for each iteration. The number of computations are more in this process. In literature such method is used many times for automatic circuit design and optimization (Fakhfakh et al., 2009; Thakker et al., 2009; Xuesong et al., 2011; Li, Yu & Li, 2008; Cooren et al., 2007). Design process is shown in Figure 1.

Figure 1.

Design process: In Loop Design and Optimization (Fakhfakh et al., 2009) (permission taken from the author for reuse of figure)

One simple case study from (Fakhfakh et al., 2009) will be helpful to understand in-loop optimization, in which amplifier circuit is designed for a given transconductance gain of an LNA. The design was having 11 design parameters so the particles declared were eleven dimensional. For a 20 particles swarm, 10000 iterations were used and solution was found in 1.57 seconds. Similar examples can be found in (Thakker et al., 2009; Xuesong et al., 2011; Li, Yu & Li, 2008; Cooren et al., 2007).