Canonization of graphs during transistor circuits decompilation

Objectives. The objective of the work is to develop the means for recognition (extraction) of high-level structures in circuits on transistor level. This allows to obtain a representation on logical level, equivalent to original flat description on transistor level. Obtaining such a representation significantly reduces the time to perform VLSI topology check, but also provides the basis for reengineering of integrated circuits and reverse engineering for detecting unauthorized attachments.
Methods. Graph based methods and software tools are proposed for recognizing topologically equivalent transistor circuits, which makes it possible to divide the set of subcircuits into topologically equivalent classes. The problem is reduced to checking the isomorphism of labeled graphs defining circuits on transistor level by canonizing them and comparing canonical labeling. The original flat and resulting two-level transistor circuits are presented in SPICE format.
Results. The proposed methods are implemented in C++ as a part of a transistor circuit decompilation program for the case without predetermined cell library. The proposed method of canonization of labeled graphs is used: to recognize topologically equivalent subcircuits among functionally equivalent subcircuits that implement logical elements; to split the set of subcircuits not recognized as logical elements into classes of topologically equivalent ones; to verify the results of extraction of the hierarchical circuit at the transistor-logic level relative to the flat circuit at the transistor level.
Conclusion. The decompilation program has been tested on practical transistor-level circuits. Experiments indicate that this tool is fast enough to process the circuits with more than one hundred thousand transistors in a few minutes on a personal computer.

1. Baker R. J. CMOS Circuit Design, Layout, and Simulation. Third ed. Wiley-IEEE Press, 2010, 1214 p.

2. Zhang N., Wunsch D. C., Harary F. The subcircuit extraction problem. Proceedings IEEE International Behavioral Modeling and Simulation Workshop, 2005, vol. 33(3), рр. 22–25.

3. Yang L., Shi C.-J. R. FROSTY: A program for fast extraction of high-level structural representation from circuit description for industrial CMOS circuits. Integration the VLSI Journal, 2006, vol. 39, no 4, рр. 311–339.

4. Cheremisinov D. I., Cheremisinova L. D. Extracting a logic gate network from a transistor-level CMOS circuit. Mikrojelektronika [Russian Microelectronics], 2019, vol. 48, no. 3, рр. 224–234. https://doi.org/10.1134/S0544126919030037 (In Russ.).

5. Abadir M. S., Ferguson J. An improved layout verification algorithm (LAVA). Proceedings of the European Design Automation Conference, Glasgow, UK, 12–15 March 1990. Glasgow, 1990, рр. 391–395.

6. Seok M. G., Park D. J., Cho G. R., Kim T. G. Framework for simulation of the Verilog/SPICE mixed model: Interoperation of Verilog and SPICE simulators using HLA/RTI for model reusability. 22nd International Conference on Very Large Scale Integration (VLSI-SoC), Playa del Garmen, Mexico, 6–8 October 2014. Playa del Garmen, 2014, рр. 1–6.

7. Kundu S. GateMaker: A transistor to gate level model extractor for simulation, automatic test pattern generation and verification. Proceedings of the International Test Conference, Washington, DC, USA, 18–23 October 1998. Washington, 1998, рр. 372–381.

8. Hunt V. D. Reengineering: Leveraging the Power of Integrated Product Development. Wiley, 1993, 283 p.

9. Rostami M., Koushanfar F., Karri R. A primer on hardware security: Models, methods, and metrics. Proceedings of the IEEE, 2014, vol. 102, no. 8, pp. 1283–1295.

10. Tehranipoor M., Koushanfar F. A survey of hardware trojan taxonomy and detection. IEEE Design & Test of Computers, 2010, vol. 27, no. 1, pp. 10–25.

11. Belous A. I., Solodukha V. A. Osnovy kiberbezopasnosti. Standarty, kontseptsii, metody i sredstva obespecheniya. Fundamentals of Cybersecurity. Standards, Concepts, Methods and Means of Support. Moscow, Tekhnosfera, 2021, 482 p. (In Russ.).

12. Cheremisinov D. I., Cheremisinova L. D. Logical gates recognition in a flat transistor circuit. Informatika [Informatics], 2021, vol. 18, no. 4, pp. 96−107. https://doi.org/10.37661/1816-0301-2021-18-4-96-107 (In Russ.).

13. Hartke S. G., Radcliffe A. J. McKay's Canonical Graph Labeling Algorithm. Communicating Mathematics, 2009, vol. 479, pp. 99–111.

14. Zakrevskij A. D., Pottosin Yu. V., Cheremisinova L. D. Logicheskiye osnovy proyektirovaniya diskretnykh ustroystv. Logical Basis for Designing Discrete Devices. Moscow, Fizmatlit, 2007, 589 p. (In Russ.).

15. Garey M. R., Johnson D. S. Computers and Intractability: A Guide to the Theory of NP-Completeness. New York, W. H. Freeman and Company, 1979, 340 р.

16. McKay B. D. Practical graph isomorphism. Congressus Numerantium, 1981, vol. 30, pp. 45–87.

17. Junttila T., Kaski P. Engineering an Efficient Canonical Labeling Tool for Large and Sparse Graphs. Proceedings Meeting on Algorithm Engineering & Expermiments SIAM, New Orleans, LA, 6 January 2007. New Orleans, 2007, pp. 135–149.

18. Cheremisinov D. I., Cheremisinova L. D. Verification of logic circuits from CMOS transistors. Novyye informatsionnyye tekhnologii v issledovanii slozhnykh struktur : materialy 13-j Mezhdunarodnoj konferencii, 7–9 sentyabrya 2020 g. [New Information Technologies in the Study of Complex Structures: Proceedings of the 13th International Conference, 7–9 September 2020]. Tomsk, Izdatel'skij dom Tomskogo gosudarstvennogo universiteta, 2020, pp. 150–151 (In Russ.).