An Improvement for The Decentralized Privacy System Using Random Linear Network Coding
Name
Saeid Mousavifar
Abstract
With the constant rise of applications, there is a huge amounts of generated sensitive and private data for each person. Hence, services need to store the generated data in a cloud or distributed hash table. Two of the main issues with external storage is privacy and security of the stored data. The privacy of such as data is preserved by implementing a permission Blockchain on top of the distributed hash table that grants the access for a user’s data to allowed services. However, the security of the data is only achieved by symmetric cryptography which is not a strong security mechanism. In this work, we apply a network coding scheme to this setup to achieve the goal of maintaining the security of the stored data. Our analysis show that by implementing random linear network coding in this setup, we achieve the security of stored data, as well as improving resiliency and retrieval time of the stored data with the expense of storage overhead and storage time. Our simulation results show that the expected retrieval time of the data is increased significantly while the expected storage time is increased with respect to the traditional setup. it also show that there is a trade-off between expected retrieval time and expected storage time in the system. These results confirm that our framework achieves the desired goal of making a faster, more resilient and secure setup for storing sensitive data with the requirement of slightly more storage size.
Graduation Thesis language
English
Graduation Thesis type
Master - Computer Science
Supervisor(s)
Rune Hylsberg Jacobsen, Satish Narayana Srirama, Ali Marandi
Defence year
2020