Optimization of Reversible Circuits Using Gate Pair Classification

Research on reversible logic gained momentum in the past decade owing to its utility in emerging areas such as quantum computing, optical computing and low power circuit implementation, etc. Reversible circuits synthesized using existing techniques often tend to be sub-optimal; thus, post-synthesis optimization techniques are usually employed to reduce the ‘circuit cost’, a metric used to compare the reversible circuits. In this paper, a set of optimization techniques is proposed to minimize the circuit cost. These techniques rely on classifying a pair of reversible gates in a given circuit based on their structural similarity. An algorithm that maps the classifications with the optimization techniques to improve the cost of a circuit is also proposed. Results obtained for a set of benchmark reversible circuits confirm that the proposed methodology performs better in terms of circuit cost when compared to those available in the literature.