Blockchain Integrated Framework for Resolving Privacy Issues in Smart City

Bedi, Pradeep; Goyal, S. B.; Kumar, Saurabh; Kumar, Shailesh

doi:10.1007/978-3-030-72139-8_6

Cited by 6 publications

(1 citation statement)

References 59 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Smart cities are rapidly emerging throughout the world, and access to a large number of Internet of things (IoT) infrastructure is necessary to address the issues in smart cities. These infrastructures transmit data all the time, share, transform, and classify data, such as medical care, insurance, transportation, and meteorology [1][2][3]. There are a lot of internal relationships hidden in these data.…”

Section: Introductionmentioning

confidence: 99%

Blockchain‐Based Privacy‐Preserving Approach Using SVML for Encrypted Smart City Data in the Era of IR 4.0

Zhang

Goyal

Tesfayohanis³

et al. 2022

Journal of Nanomaterials

Self Cite

View full text Add to dashboard Cite

The collection of amounts of useful data from various IoT devices through machine learning (ML) techniques has been extensively applied in lots of areas of the smart city. To improve the efficiency of machine learning, people realize efficient data classification by supporting the ML model, namely, the support vector machine (SVM) model, which is used in the realization of linear classification. However, data security and data protection are not addressed in SVM classifier training from multiple entity-labeled IoT data. In the existing solution, they depend on the implicit hypothesis that learning data can be dependable collected from complex data providers, which is often not the reality. To solve the question between the above problems, in this article, we put forward a secure support vector machine—a secure SECURE SVM training scheme for protecting the privacy on encrypted IoT data based on blockchain. Firstly, a secure and dependable data encryption sharing platform is established among complex data providers by using blockchain technology, which is recorded in a distributed ledger. Secondly, the homomorphic cryptosystem is used to design secure components, and the secure polynomial multiplication is improved to design a secure SVM training algorithm, which only needs two interactions in one iteration and does not need a trusted third party. Finally, the experimental results display that the plan ensures the confidentiality of sensitive data of each data provider and the confidentiality and validity of SVM model parameters of data analysts.

show abstract

Section: Introductionmentioning

confidence: 99%