Safeguarding Privacy by utilizing SC-DℓDA Algorithm in Cloud-Enabled Multi Party Computation

• D. Dhinakaran,
• G. Prabaharan,
• K. Valarmathi,
• S.M. Udhaya Sankar,
• R. Sugumar,

Vol. 19, No. 2, February 28, 2025
10.3837/tiis.2025.02.014, Download Paper (Free):

• Privacy Preserving
• Spectrasave Encryption
• Dynamic ℓ-Diversity Anonymity
• Security
• Multi Party Computation

Abstract

Privacy Preserving data processing for healthcare has been a high concern of late as the volume of sensitive medical data that is stored and shared across platforms continues to grow. Data utility for analysis is a great challenge especially in cloud-based health care systems, as well as protecting the privacy of the individual. Based on that, this paper proposes a novel privacy preserving health care data framework that incorporates Spectrasave Encryption (SE) and Dynamic ℓ-Diversity Anonymity (DℓDA) to handle this problem efficiently. We encrypt sensitive data to the healthcare users using Spectrasave encryption before it is sent to a trusted cloud server. The DℓDA algorithm is applied on the cloud server, which decrypts and anonymizes the data under privacy principles of both k-anonymity and ℓ-diversity, but with minimum re-identification risks. Furthermore, the cloud server kens weighted t-closeness to make the cloud server more secure, whereby sensitive distributions of attributes are still sufficiently obfuscated. Our results demonstrate that SE-DℓDA outperforms a number of existing approaches on various measures. For instance, SE-DℓDA reduces the computational cost of encryption and decryption by 15.1% and 16.4%, respectively. Our proposed approach reduces communication overhead by up to 28.3%. In particular, SE-DℓDA is more scalable with larger data set sizes and also with a larger number of data owners, and in fact manages to perform better as a function of the number of data owners between 2 and 10, with 6.8% better performance than the competing methods.

Statistics

Cite this article