Abstract:
Research has been done on the metaheuristic optimization of RSFQ circuits, but these are typically limited to single criteria (e.g., yield). The importance of using multiple criteria and the relevance of each of the criteria is discussed. The application of data envelopment analysis (DEA) and an interactive normalization technique as multi-criteria evaluation methods for the optimization of RSFQ circuits are investigated. We find that these evaluation methods each have clear advantages and disadvantages over each other. The normalization technique requires the analysis designer to have a good understanding of circuit design to be accurate, while theDEAmethod may yield many results that are ranked equally efficient. © 2010 IEEE.