On Blockchain and IoT Integration Platforms: Current Implementation Challenges and Future Perspectives

Table 4

Blockchain-based machine learning privacy solutions for IoT devices and networks.

Ref.	Attack	Use case	Metric	Algorithm	Blockchain used	Description
Shen et al. [114]	Data privacy (ciphertext model, known as the background model)	Smart cities	Accuracy	SVM (secureSVM)	—	A secure SVM training algorithm in multipart scenarios was created on IoT data. Proof-of-work consensus algorithm was used. Accuracy: 93.89%
Mendis et al. [108]	Data leakage	Multipurpose/general	Accuracy	CNN	Ethereum	Proof-of-stake consensus algorithm used. This research looked at training machine learning models in a distributed fashion on multiple Raspberry Pi 3’s.
Arachchige et al. [115]	Data privacy	Industrial IoT		Federated ML	Ethereum	A framework was introduced called PriModChain that prevents leak of sensitive data from IoTs to adversarial networks.