ENHANCING CYBERSECURITY AND DATA PROTECTION IN INFORMATION SYSTEMS BASED ON PROOF OF FRIENDSHIP AND ZERO-KNOWLEDGE PROOF CIRCUITS BUILT ON THE SM3 HASH FUNCTION

Abstract

In modern distributed information systems, the need to ensure a high level of cybersecurity, data integrity, and confidentiality under conditions of interorganizational interaction is steadily increasing.Blockchain technologies enhance transparency and trust among participants; however, traditional consensus mechanisms are accompanied by significant computational overhead, risks of centralization, and limited capabilities for protecting sensitive information.These issues are particularly acute in corporate environments of small and medium-sized enterprises, where the computational resources of network nodes are constrained while the requirements for business data confidentiality remain high.A promising direction is the integration of Zero-Knowledge Proof (ZKP) mechanisms, which enable verification of operation correctness without disclosing the underlying data.Nevertheless, their practical adoption is hindered by the high cost of proof construction for classical cryptographic primitives.In particular, for the SM3 hash function there are no efficient optimized implementations of preimage proofs, and its bit-oriented structure leads to a substantial increase in circuit size and proof generation time, making its use infeasible in resource-constrained environments.This paper proposes a dockerized private blockchain architecture oriented toward corporate environments with limited resources, combining the trust-oriented Proof of Friendship consensus with Zero-Knowledge Proof mechanisms.The key result is the development of an approach for optimizing SM3 hash preimage proofs in ZKP systems.The paper introduces principles of manual optimization of the SM3 circuit representation, including reduction of bitwise operations, aggregation of 1965 constraints, optimization of message expansion, and reduction of round depth.It is shown that these transformations significantly decrease the size of arithmetic circuits and proof generation time compared to naive algorithm translation, enabling practical use of SM3 in zero-knowledge systems and corporate blockchain solutions.The proposed approach provides a balance between blockchain transparency and business data confidentiality, forming a "trust but do not disclose" model.The obtained results establish a scientific and practical foundation for deploying privacypreserving computation in distributed information systems and for developing nextgeneration secure blockchain platforms.