Using consensusless covers for fast operating on Boolean functions
The paper presents a method for fast operating on covers of Boolean functions. The method develops the one based on the unate paradigm (UP). The proposed method differs from the UP one in two aspects (1) the initial cover is decomposed into a set of prime rather than unate subcovers, (2) prime covers are obtained by applying to each reched not prime subcover either branching by the Shannon expansion or a procedure of making the subcover consensusless in a variable by the consensus operation. Experiments on MCNC-91 two-level logic benchmarks and random functions show that operations based on the proposed method are less laborious than their UP based counterparts.
Unable to display preview. Download preview PDF.