A security protection method and system for generating honey spots in smart contracts based on large models

By generating honeypots in smart contracts using large models, and utilizing multimodal data training and adversarial training to generate contracts with hidden vulnerabilities, combined with real-time monitoring and tiered response, this approach addresses the shortcomings of traditional smart contract protection methods and achieves efficient and secure smart contract protection.

CN121808769BActive Publication Date: 2026-06-30GUANGZHOU UNIVERSITY
View PDF 4 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
GUANGZHOU UNIVERSITY
Filing Date
2026-03-06
Publication Date
2026-06-30

AI Technical Summary

Technical Problem

Traditional smart contract security protection methods rely on static analysis and manual auditing, which makes it difficult to detect new dynamic attacks. Furthermore, the generation of honey spots lacks intelligence and concealment, posing a risk to information storage.

Method used

The method of generating honey spots in smart contracts using large models generates contracts with hidden vulnerabilities through multimodal data training and adversarial training. These contracts are monitored in real time and stored on the blockchain. Combined with anomaly detection and tiered response measures, dynamic defense is achieved.

Benefits of technology

It enhances the concealment and defense effectiveness of honeypots, ensures information security and reliability, and enables timely identification and efficient response to malicious behavior.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121808769B_ABST
    Figure CN121808769B_ABST
Patent Text Reader

Abstract

This invention discloses a security protection method and system for generating smart contract honeypots based on a large model, comprising: collecting and preprocessing multimodal training data containing vulnerability features, behavioral features, and semantic features; performing adversarial training on the large model based on the dataset to generate smart contract honeypots that appear to function normally but contain hidden vulnerabilities; storing the honeypot contract information on the blockchain after encapsulation and digital signature, and deploying it to the blockchain network; monitoring blockchain transaction logs in real time, identifying honeypot calls and extracting features for anomaly detection; determining malicious users based on anomaly scoring combined with preset rules, and implementing tiered response measures such as freezing on-chain assets, adding to a blacklist, and linking to internal network isolation. This invention utilizes a large model to dynamically generate highly concealed honeypots, combined with blockchain trusted storage and real-time monitoring, to achieve proactive, precise, and collaborative protection against smart contract attacks, effectively improving the security defense capabilities of the blockchain ecosystem.
Need to check novelty before this filing date? Find Prior Art