Principal component analysis (PCA) is an essential algorithm for dimensionality reduction in many data science domains. We address the problem of performing a federated PCA on private data distributed among multiple data providers while ensuring data confidentiality. Our solution, SF-PCA, is an end-to-end secure system that preserves the confidentiality of both the original data and all intermediate results in a passive-adversary model with up to all-but-one colluding parties. SF-PCA jointly leverages multiparty homomorphic encryption, interactive protocols, and edge computing to efficiently interleave computations on local cleartext data with operations on collectively encrypted data. SF-PCA obtains results as accurate as non-secure centralized solutions, independently of the data distribution among the parties. It scales linearly or better with the dataset dimensions and with the number of data providers. SF-PCA is more precise than existing approaches that approximate the solution by combining local analysis results, and between 3x and 250x faster than privacy-preserving alternatives based solely on secure multiparty computation or homomorphic encryption. Our work demonstrates the practical applicability of secure and federated PCA on private distributed datasets.

Abstract In this paper, we address the problem of privacy-preserving distributed learning and the evaluation of machine-learning models by analyzing it in the widespread MapReduce abstraction that we extend with privacy constraints. We designspindle(Scalable Privacy-preservINg Distributed LEarning), the first distributed and privacy-preserving system that covers the complete ML workflow by enabling the execution of a cooperative gradient-descent and the evaluation of the obtained model and by preserving data and model confidentiality in a passive-adversary model with up to N −1 colluding parties.spindleuses multiparty homomorphic encryption to execute parallel high-depth computations on encrypted data without significant overhead. We instantiatespindlefor the training and evaluation of generalized linear models on distributed datasets and show that it is able to accurately (on par with non-secure centrally-trained models) and efficiently (due to a multi-level parallelization of the computations) train models that require a high number of iterations on large input data with thousands of features, distributed among hundreds of data providers. For instance, it trains a logistic-regression model on a dataset of one million samples with 32 features distributed among 160 data providers in less than three minutes.

Data sharing has become of primary importance in many domains such as big-data analytics, economics and medical research, but remains difficult to achieve when the data are sensitive. In fact, sharing personal information requires individuals’ unconditional consent or is often simply forbidden for privacy and security reasons. In this paper, we propose Drynx, a decentralized system for privacy-conscious statistical analysis on distributed datasets. Drynx relies on a set of computing nodes to enable the computation of statistics such as standard deviation or extrema, and the training and evaluation of machine-learning models on sensitive and distributed data. To ensure data confidentiality and the privacy of the data providers, Drynx combines interactive protocols, homomorphic encryption, zero-knowledge proofs of correctness, and differential privacy. It enables an efficient and decentralized verification of the input data and of all the system’s computations thus provides auditability in a strong adversarial model in which no entity has to be individually trusted. Drynx is highly modular, dynamic and parallelizable. Our evaluation shows that it enables the training of a logistic regression model on a dataset (12 features and 600,000 records) distributed among 12 data providers in less than 2 seconds. The computations are distributed among 6 computing nodes, and Drynx enables the verification of the query execution’s correctness in less than 22 seconds.

Proofs of Retrievability (POR) are cryptographic proofs which provide assurance to a single tenant (who creates tags using his secret material) that his files can be retrieved in their entirety. However, POR schemes completely ignore storage-efficiency concepts, such as multi-tenancy and data deduplication, which are being widely utilized by existing cloud storage providers. Namely, in deduplicated storage systems, existing POR schemes would incur an additional overhead for storing tenants' tags which grows linearly with the number of users deduplicating the same file. This overhead clearly reduces the (economic) incentives of cloud providers to integrate existing POR/PDP solutions in their offerings. In this paper, we propose a novel storage-efficient POR, dubbed SPORT, which transparently supports multi-tenancy and data deduplication. More specifically, SPORT enables tenants to securely share the same POR tags in order to verify the integrity of their deduplicated files. By doing so, SPORT considerably reduces the storage overhead borne by cloud providers when storing the tags of different tenants deduplicating the same content. We show that SPORT resists against malicious tenants/cloud providers (and against collusion among a subset of the tenants and the cloud). Finally, we implement a prototype based on SPORT, and evaluate its performance in a realistic cloud setting. Our evaluation results show that our proposal incurs tolerable computational overhead on the tenants and the cloud provider.

Abstract Current solutions for privacy-preserving data sharing among multiple parties either depend on a centralized authority that must be trusted and provides only weakest-link security (e.g., the entity that manages private/secret cryptographic keys), or leverage on decentralized but impractical approaches (e.g., secure multi-party computation). When the data to be shared are of a sensitive nature and the number of data providers is high, these solutions are not appropriate. Therefore, we present UnLynx, a new decentralized system for efficient privacy-preserving data sharing. We considermservers that constitute a collective authority whose goal is to verifiably compute on data sent fromndata providers. UnLynxguarantees the confidentiality, unlinkability between data providers and their data, privacy of the end result and the correctness of computations by the servers. Furthermore, to support differentially private queries, UnLynxcan collectively add noise under encryption. All of this is achieved through a combination of a set of new distributed and secure protocols that are based on homomorphic cryptography, verifiable shuffling and zero-knowledge proofs. UnLynxis highly parallelizable and modular by design as it enables multiple security/privacy vs. runtime tradeoffs. Our evaluation shows that UnLynxcan execute a secure survey on 400,000 personal data records containing 5 encrypted attributes, distributed over 20 independent databases, for a total of 2,000,000 ciphertexts, in 24 minutes.