Interior Search Algorithm integrated Matlab-SPICE Interface for Optimization of CMOS Inverter Switching Characteristics

Interior Search Algorithm (ISA) has been utilized in this paper to realize a CMOS inverter with symmetric and fast switching characteristics. Interior search algorithm, a nature based metaheuristic algorithm uses two distinct search methods - mirror search and composition optimization. This algorithm is used to achieve fast and symmetric switching characteristics in the transient response of the optimized inverter by computing the transistor sizes and capacitor value. The results obtained from the Interior Search Algorithm optimized fast symmetric inverter are compared to previously used optimization algorithms like Particle Swarm Optimization (PSO), Real-coded Genetic Algorithm (RGA), Cuckoo Search Algorithm (CSA) and Particle Swarm Optimization with Constriction-Factor and Inertia-Weight Approach (PSO-CFIWA). A Matlab-SPICE interface was developed using Tool Command Language (Tcl) as an enhancement over the standalone Matlab approach. SPICE simulations were used for iterative optimization of the fitness function in the interface. The proposed approach led to 94-99% improvement in SPICE results as compared to other standalone Matlab optimization approaches.