Logical effort base adder circuits transistor sizing using constriction factor and mutative variants of particle swarm optimization algorithm / Muhammad Aiman Johari

In Semiconductor world, the design and fabrication of Integrated Circuit (IC) associated with development time, operating speed and power requirements. The goodness of each design must be evaluated before it is chosen, especially on speed of the circuits where it represent the time taken to execute a specific function or most commonly known as delay. Conventional methods use repetitive manual testing guided by Logical Effort (LE). LE provides an easy way to compare and select circuit topologies, choose the best number of stages for path and estimate path delay. The proposal of Particle Swarm Optimization (PSO) with constriction factor (PSO-CF) and mutative variants (PSO-M) presented in this thesis attempts to create an automated process of transistor sizing optimization. The method attempts to get the target circuit delay on tested circuit's critical path based on LE calculation that accepts generated transistor size by both PSO variants as inputs to fitness function. The optimization of the transistor size will stop if maximum iteration reached of different between PSO's found delay and objective delay is very small (near or similar to c0'). Various parameters, such as swarm size and iterations were tested under different initial positions to verify PSO's performance on a adder circuits namely modified half-adder (M-HA), modified full-adder (M-FA) and modified ripple-carry adder (MRCA). The experiments reported in this thesis showed that both PSO variants were efficient to automatically find the optimum transistor size with solution range of [1(T2,1(T15] for PSO-CF and [10°, 1(T16] for PSO-M.