Securing instant messaging based on blockchain with machine learning

【Author】 Yi, Haibo

【Source】SAFETY SCIENCE

【影响因子】6.392

【Abstract】Instant Messaging (IM) offers real-time communications between two or more participants on Internet. Nowadays, most Ms take place on mobile applications, such as WhatsApp, WeChat, Viber and Facebook Messenger, which have more users than social networks, such as Twitter and Facebook. Among the applications of IMs, online shopping has become a part of our everyday life, primarily those who are busiest. However, transaction disputes are often occurred online shopping. Since most IMs are centralized and message history is not stored in the center, the messaging between users and owners of online shops are not reliable and traceable. In China, online shopping sales have soared from practically zero in 2003 to nearly 600 hundred million dollars last year, and now top those in the United States. It is very crucial to secure the instant messaging in online shopping in China. We present techniques to exploit blockchain and machine learning algorithms to secure instant messaging. Since the cryptography of Chinese national standard is encouraged to adopt in security applications of China, we propose a blockchain-based IM scheme with the Chinese cryptographic bases. First, we design a message authentication model based on SM2 to avoid the counterfeit attack and replay attack. Second, we design a cryptographic hash mode based on SM3 to verify the integrity of message. Third, we design a message encryption model based on SM4 to protect the privacy of users. Besides, we propose a method based on machine learning algorithms to monitor the activity on blockchain to detect anomaly. To prove and verify the blockchain-based IM scheme, a blockchain-based IM system has been designed on Linux platforms. The implementation result shows that it is a practical and secure IM system, which can be applied to a variety of instant messaging applications directly.

【Keywords】Instant Messaging (IM); BlockChain; Machine learning; Distributed Ledger Technology (DLT); Safety and security

【发表时间】2019 DEC

【收录时间】2022-01-02

【文献类型】

【主题类别】

--