计算机科学
智能合约
块链
计算机安全
脆弱性(计算)
人工智能
模糊逻辑
机器学习
数据挖掘
作者
Daojing He,Rui Wu,Xinji Li,Sammy Chan,Mohsen Guizani
出处
期刊:IEEE Internet of Things Journal
[Institute of Electrical and Electronics Engineers]
日期:2023-07-15
卷期号:10 (14): 12178-12185
被引量:12
标识
DOI:10.1109/jiot.2023.3241544
摘要
With the wide application of Internet of Things and blockchain, research on smart contracts has received increased attention, and security threat detection for smart contracts is one of the main focuses. This article first introduces the common security vulnerabilities in blockchain smart contracts, and then classifies the vulnerabilities detection tools for smart contracts into six categories according to the different detection methods: 1) formal verification method; 2) symbol execution method; 3) fuzzy testing method; 4) intermediate representation method; 5) stain analysis method; and 6) deep learning method. We test 27 detection tools and analyze them from several perspectives, including the capability of detecting a smart contract version. Finally, it is concluded that most of the current vulnerability detection tools can only detect vulnerabilities in a single and old version of smart contracts. Although the deep learning method detects fewer types of smart contract vulnerabilities, it has higher detection accuracy and efficiency. Therefore, the combination of static detection methods, such as deep learning method and dynamic detection methods, including the fuzzy testing method to detect more types of vulnerabilities in multi-version smart contracts to achieve higher accuracy is a direction worthy of research in the future.
科研通智能强力驱动
Strongly Powered by AbleSci AI