Home         Authors   Papers   Year of conference   Themes   Organizations        To MES conference

On complexity of inverter graphs for Boolean functions of small number of variables  

Authors
 Lozhkin S.A.
 Zizov V.S.
 Shupletsov M.S.
 Zhukov V.V.
 Khzmalian D.E.
 Belyankov O.O.
Date of publication
 2020
DOI
 10.31114/2078-7707-2020-4-95-102

Abstract
 In this paper, we consider the problem of exact synthesis inverter graphs for Boolean functions of small number of variables. Main task was to construct catalogue of such schemes. This problem is considered for two models of inverter graphs: and-inverter graph and majority-inverter graph. Algorithms for exact synthesis and corresponding software have been developed to solve this problem. These methods include re-writing, SAT-based synthesis, graph enumeration, iterative minimization and can be applied to any inverter graph model. The qualitative analysis of these methods had been performed. Near-optimal inverter graphs for all functions of five or less variables were obtained. We have all optimal majority-inverter graph realizations with complexity equal or less than 4. An upper bound of the complexity for all functions of five variables was established 11 for the MIG model and 17 for the AIG model. The average complexity of majority-inverter graphs in our database is 7.701 and the average complexity of and-inverter graphs is 12.800. As a result, average complexity for MIG model was shown to be approximately 39.8% less than the average complexity of AIG model. An implementation in the MIG model was demonstrated for the single worst-case equivalence class of Boolean functions (Fig. 8).
Keywords
 methods for exact synthesis, Boolean logic network, Post basis, complexity of Boolean functions, average complexity, majority function, circuit design database.
Library reference
 Lozhkin S.A., Zizov V.S., Shupletsov M.S., Zhukov V.V., Khzmalian D.E., Belyankov O.O. On complexity of inverter graphs for Boolean functions of small number of variables // Problems of Perspective Micro- and Nanoelectronic Systems Development - 2020. Issue 4. P. 95-102. doi:10.31114/2078-7707-2020-4-95-102
URL of paper
 http://www.mes-conference.ru/data/year2020/pdf/D103.pdf

Copyright © 2009-2024 IPPM RAS. All Rights Reserved.

Design of site: IPPM RAS