A Heuristic-Driven and Cost Effective Majority/ Minority Logic Synthesis for Post-CMOS Emerging Technology

Abstract

Due to the physical restriction of current CMOS technology, emerging technologies that have majority logic gate as a base component are being explored. The process of transforming from boolean network to the majority logic network is called majority logic synthesis (MLS). Hence, the contributions of this work is as follows: (i) a novel heuristic-driven tabular approach for majority logic synthesis has been presented that overcomes the scalability problem of previous synthesis algorithms, (ii) a heuristic named as matching difference (MD) is proposed to guide the synthesis process, (iii) to maintain the trade-off between area and delay during majority logic synthesis a novel criterion “cost of circuit”(CoC) has been proposed, (iv) For reduction of majority circuit delay during MLS, an extended library based on five-input and three-input majority gates is presented, and (v) a post-synthesis optimization method is proposed based on majority algebra. Based on experiments with MCNC benchmarks, it is verified that the proposed approach accomplishes an average diminution of 24% at the majority level and 31% in the majority gate count. Further, while executing a case study with quantum dot cellular automata(QCA), the proposed methodology is able to achieve an average diminution of 36% in delay and 15% in circuit cost with a penalty of 2% in the area.

Keywords

• Majority gate
• QCA
• NML
• majority logic synthesis