Many Boolean functions arising in circuit applications exhibit structural regularities that can be exploited to improve logic synthesis. Among them, autosymmetry is an XOR-based property that allows a function to be studied in a lower-dimensional space through a suitable restriction. In this work, we propose an efficient incremental heuristic for testing autosymmetry directly from And-Inverter Graph representations, without requiring the preliminary construction of a BDD. The method can identify the autosymmetry degree when run to completion, or return a certified lower bound if interrupted, while exploiting algebraic properties of the output to improve performance. Experimental results show the effectiveness of the proposed approach.
Incremental Autosymmetry Testing of Regular Functions through AND-Inverter Graphs / G. Cuciniello, A.B.. - In: IEEE EMBEDDED SYSTEMS LETTERS. - ISSN 1943-0663. - (2026). [Epub ahead of print] [10.1109/LES.2026.3728203]
Incremental Autosymmetry Testing of Regular Functions through AND-Inverter Graphs
G. CucinielloPrimo
;V. CirianiUltimo
2026
Abstract
Many Boolean functions arising in circuit applications exhibit structural regularities that can be exploited to improve logic synthesis. Among them, autosymmetry is an XOR-based property that allows a function to be studied in a lower-dimensional space through a suitable restriction. In this work, we propose an efficient incremental heuristic for testing autosymmetry directly from And-Inverter Graph representations, without requiring the preliminary construction of a BDD. The method can identify the autosymmetry degree when run to completion, or return a certified lower bound if interrupted, while exploiting algebraic properties of the output to improve performance. Experimental results show the effectiveness of the proposed approach.| File | Dimensione | Formato | |
|---|---|---|---|
|
Incremental_Autosymmetry_Testing_of_Regular_Functions_through_AND-Inverter_Graphs.pdf
accesso aperto
Tipologia:
Post-print, accepted manuscript ecc. (versione accettata dall'editore)
Licenza:
Creative commons
Dimensione
230.15 kB
Formato
Adobe PDF
|
230.15 kB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.




