Transfert inconscient | EPFL Graph Search

Le transfert inconscient (oblivious transfer, en anglais) est une primitive cryptographique où un expéditeur transmet une information, sélectionnée parmi plusieurs envois possibles, à un destinataire, sans que l'expéditeur puisse connaître le choix du destinataire, ni que le destinataire puisse connaître les informations qu'il n'a pas demandées. Par exemple, Wikipédia propose plusieurs articles ; avec le transfert inconscient, un utilisateur peut demander à consulter un article sans que Wikipédia puisse savoir quel article a été consulté. Cette primitive a été introduite en 1981 par Michael Rabin, dans un manuscrit intitulé How to exchange secrets with oblivious transfer. Dans la version de Rabin, basée sur le chiffrement RSA, l’expéditeur transmet un message que le destinataire reçoit avec une probabilité de 1/2, sans que l'expéditeur puisse savoir si la réception a eu lieu ou non. En 1985, les travaux de Even, Goldreich et Lempel ont proposé une version améliorée du transfert inc

The Limits of Composable Crypto with Transferable Setup Devices

Miyako Ohkubo, Serge Vaudenay

UC security realized with setup devices imposes that single instances of these setups are used. In most cases, UC-realization relies further on other properties of the setups devices, like tamper-resistance. But what happens in stronger versions of the UC framework, like EUC or JUC, where multiple instances of these setups are allowed? Can we formalise what it is about setups like these which makes them sometimes hinder UC, JUC, EUC realizability? In this paper, we answer this question. As such, we formally introduce transferable setups, which can be viewed as setup devices that do not (publicly) disclose if they have been maliciously passed on. Further, we prove the general result that one cannot realize oblivious transfer (OT) or any "interesting" 2-party protocol using transferable setups in the EUC model. As a by-product, we show that physically unclonable functions (PUFs) themselves are transferable devices, which means that one cannot use PUFs as a global setups; this is interesting because non-transferability is a weaker requirement than locality, which until now was the property informally blamed for UC-impossibility results regarding PUFs as global setups. If setups are transferable (i.e., they can be passed on from one party to another without explicit disclosure of a malicious transfer), then they will not intrinsically leak if a relay attack takes place. Indeed, we further prove that if relay attacks are possible then oblivious transfer cannot be realized in the JUC model. Linked to the prevention of relaying, authenticated channels have historically been an essential building stone of the UC model. Related to this, we show how to strengthen some existing protocols UC-realized with PUFs, and render them not only UC-secure but also JUC-secure.

ACM2015

Privacy-Preserving Federated Analytics using Multiparty Homomorphic Encryption

David Jules Froelicher

Analyzing and processing data that are siloed and dispersed among multiple distrustful stakeholders is difficult and can even become impossible when the data are sensitive or confidential. Current data-protection and privacy regulations highly restrict the sharing and outsourcing of personal information among stakeholders that are in different jurisdictions. Sharing data is, however, required in many domains. The medical sector is a paradigmatic example: Privacy is paramount and data sharing is needed in numerous applications where data is scarce and scattered among multiple stakeholders around the world. Existing privacy-preserving solutions for federated analytics (FA) rely either (1) on data centralization or outsourcing to a limited number of entities, which incur multiple security and trust issues, or (2) on the exchange of cleartext aggregated and optionally obfuscated data, which can leak personal information or introduce bias in the final result. In this thesis, our goal is (1) to propose privacy-preserving federated solutions for exploration, and for statistical and machine-learning analyses on data held by multiple distrustful stakeholders, and (2) to analyze and evaluate the proposed systems, thus showing that they provide an efficient, secure, scalable, and accurate alternative to existing solutions for FA by proving their utility in real-world state-of-the-art biomedical studies. We rely on multiparty homomorphic encryption (MHE). MHE combines secure multiparty computation (SMC) techniques with homomorphic encryption (HE) by pooling the advantages of both SMC and HE, i.e., interactivity and flexibility, and by minimizing their disadvantages, i.e., difficulty in scaling to a large number of parties and computation complexity.First, we design UnLynx, a system that enables privacy-preserving federated data exploration on a distributed dataset held by multiple data-providers (DPs), where N-1 out of N of the nodes performing the computations can be malicious. We build interactive protocols by relying on ElGamal additive homomorphic encryption (AHE) and ensure that each untrusted-node operation can be publicly verified by means of zero-knowledge proofs (ZKPs). We then explore how statistics, e.g., standard deviation and variance, can be computed by relying on AHE and ZKPs through the design of another system named Drynx. In Drynx, we also explore how to limit the influence of an entity that inputs wrong data in the system, and we propose an efficient federated solution for correctness verification.We propose Spindle a solution for secure cooperative gradient descent on federated data that we instantiate for the privacy-preserving training and oblivious evaluation of generalized linear models. Spindle covers the entire machine-learning workflow, as it enables oblivious predictions to be performed on a trained model that remains secret. It ensures both data and model confidentiality in a passive adversarial model in which N-1 out of N DPs can collude.Finally, we demonstrate that the solutions proposed in this thesis can be efficient enablers for large-scale, sensitive, multi-site biomedical studies. We design and test, by replicating recent medical studies, secure workflows for the federated execution of computations that span from analyses with low computational complexity, such as survival analyses, to analyses with high computational complexity such as genome-wide association studies on millions of variants.

EPFL2021

"Once Upon a Place": Compute Your Meeting Location Privately

Imad Aad, Igor Bilogrevic, Jean-Pierre Hubaux, Murtuza Jadliwala

Popular services such as Doodle Mobile and Tymelie are extremely useful planning tools that enable mobile-phone users to determine common meeting time(s) for events. Similar planning tools for determining optimal meeting locations, based on the location preferences of the users, are highly desirable for event planning and management in popular mobile phone applications, such as taxi sharing, route planning and mobile participatory sensing. Yet, they have received very little attention by researchers. An important, and often overlooked, facet of such planning applications is the privacy of the participating users and their preferences; users want to agree on a meeting location without necessarily revealing their location preferences to the service provider or to the other users. In this paper, we address the problem of privacy-preserving optimal meeting-location computation, especially focusing on its applicability to current mobile devices and applications. We first define the notion of privacy in such computations. Second, we model the problem of optimal meeting-location computation as a privacy-preserving k-center problem and we design two solutions; both solutions take advantage of the homomorphic properties of well-known cryptosystems by Boneh-Goh-Nissim, ElGamal and Paillier in order to perform oblivious computations. Third, we implement the proposed solutions on a testbed of the latest generation Nokia mobile devices and study their performance. Finally, we assess the utility and expectations, in terms of privacy and usability, of the proposed solutions by means of a targeted survey and user-study of mobile-phone users.

2011

Privacy-Preserving Federated Analytics using Multiparty Homomorphic Encryption

David Jules Froelicher

EPFL2021