Open-source Windows-native application for file synchronization with secure data deduplication

Data deduplication is used to optimize storage consumption by reducing redundancy. Besides the cost advantages, data deduplication also introduce security problems, such as content learning attacks or data leaks. In one of our previous contributions, we proposed a data deduplication implementation. In this paper, we introduce an improved version of our previous work, focusing on performance and user experience. The proposed implementation is publicly available on Github and uses techniques such as convergent encryption, Merkle trees and on-premise deduplication gateway. In terms of performance, uploads are 3 times faster for small documents (up to 10Mb) and more 10 times faster for large documents (over 400Mb). In terms of user experience, the application is now seamlessly integrated with Windows and users do not need to use a custom application to benefit from data deduplication. The proposed design is able to secure documents that are deduplicated and stored in a public cloud.