Federated Learning (FL) ensures collaborative learning among multiple clients while maintaining data locally. However, the traditional synchronous FL solutions have lower accuracy and require more ...communication time in scenarios where most devices drop out during learning. Therefore, we propose an Asy nchronous F ederated L earning (AsyFL) scheme using time-weighted and stale model aggregation, which effectively solves the problem of poor model performance due to the heterogeneity of devices. Then, we integrate Symmetric Homomorphic Encryption (SHE) into AsyFL to propose Asy nchronous P rivacy- P reserving F ederated L earning (Asy-PPFL), which protects the privacy of clients and achieves lightweight computing. Privacy analysis shows that Asy-PPFL is indistinguishable under Known Plaintext Attack (KPA) and convergence analysis proves the effectiveness of our schemes. A large number of experiments show that AsyFL and Asy-PPFL can achieve the highest accuracy of 58.40% and 58.26% on Cifar-10 dataset when most clients (i.e., 80%) are offline or delayed, respectively.
Fully Homomorphic Encryption (FHE) allows outsourced computation on clients' encrypted data while preserving data privacy. FHE's high computational intensity incurs high overhead from data transfer ...with hardware such as CPU, GPU, and FPGA, due to the inherent separation between computing and data. To overcome this limitation, Compute-Enabled RAM (CE-RAM) has been explored; however, prior work using CE-RAM to accelerate FHE only explores a simple implementation of a finite-field FHE scheme and did not explore algorithmic optimizations.
In this paper, we investigate CE-RAM acceleration FHE more deeply, implementing both the finite-field B/FV and torus-based TFHE cryptosystems in CE-RAM with common FHE optimizations. This is the first work to explore using CE-RAM to accelerate TFHE. For B/FV, we explore parameter-specific algorithmic optimizations specifically designed for CE-RAM friendliness. We evaluate our implementation as compared to prior work in CE-RAM FHE acceleration and other hardware acceleration strategies. We demonstrate speedups of up to 784x for B/FV homomorphic multiplication and 38x for TFHE bootstrapping as compared to CPU implementations. We also discuss the overhead of CE-RAM for FHE on energy and area consumption, showing comparable or improved performance as compared to other work or hypothetical near-memory accelerators.
Aimed at the existing problems, such as the improving requirements on privacy, and the low quality of collected data in mobile crowd sensing, one online Quality-based Privacy-preserving Task ...Allocation (QPTA) mechanism is considered in this paper. First, a travel budget-related polynomial is designed to measure the influence of candidate tasks on the whole task quality, and an online task quality-based allocation model is designed to maximize the total task quality when both the budget and time constraints can be satisfied simultaneously. Second, when the perceived data of task participants and requesters are encrypted values, with the introduction of typical Paillier homomorphic encryption technology, the division, the square root, and the Euclidean distance calculations of the encrypted values are then designed. In this way, the encrypt calculations of the task allocation algorithm are solved, and the data privacy of the task allocation system is thus preserved. Finally, the bilateral privacy analysis of the proposed method is illustrated, and some simulations are carried out to show the effectiveness of the proposed method. Simulation results show that the proposed method can allocate tasks effectively while guaranteeing the privacy of both the participants and requesters. It outperforms the most related methods in terms of average perception quality.
We present a novel approach to fully homomorphic encryption (FHE) that dramatically improves performance and bases security on weaker assumptions. A central conceptual contribution in our work is a ...new way of constructing leveled, fully homomorphic encryption schemes (capable of evaluating arbitrary polynomial-size circuits of a-priori bounded depth), without Gentry’s bootstrapping procedure.
Specifically, we offer a choice of FHE schemes based on the learning with error (LWE) or Ring LWE (RLWE) problems that have 2
λ
security against known attacks. We construct the following.
(1) A leveled FHE scheme that can evaluate depth-
L
arithmetic circuits (composed of fan-in 2 gates) using
O
(
λ
.
L
3) per-gate computation, quasilinear in the security parameter. Security is based on RLWE for an approximation factor exponential in
L
. This construction does not use the bootstrapping procedure.
(2) A leveled FHE scheme that can evaluate depth-
L
arithmetic circuits (composed of fan-in 2 gates) using
O
(
λ
2) per-gate computation, which is independent of
L
. Security is based on RLWE for quasipolynomial factors. This construction uses bootstrapping as an optimization.
We obtain similar results for LWE, but with worse performance. All previous (leveled) FHE schemes required a per-gate computation of
Ω
(
λ
3.5), and all of them relied on subexponential hardness assumptions.
We introduce a number of further optimizations to our scheme based on the Ring LWE assumption. As an example, for circuits of large width (e.g., where a constant fraction of levels have width
Ω
(
λ
)), we can reduce the per-gate computation of the bootstrapped version to
O
(
λ
), independent of
L
, by batching the bootstrapping operation. At the core of our construction is a new approach for managing the noise in lattice-based ciphertexts, significantly extending the techniques of Brakerski and Vaikuntanathan 2011b.
We present a privacy-preserving deep learning system in which many learning participants perform neural network-based deep learning over a combined dataset of all, without revealing the participants' ...local data to a central server. To that end, we revisit the previous work by Shokri and Shmatikov (ACM CCS 2015) and show that, with their method, local data information may be leaked to an honest-but-curious server. We then fix that problem by building an enhanced system with the following properties: 1) no information is leaked to the server and 2) accuracy is kept intact, compared with that of the ordinary deep learning system also over the combined dataset. Our system bridges deep learning and cryptography: we utilize asynchronous stochastic gradient descent as applied to neural networks, in combination with additively homomorphic encryption. We show that our usage of encryption adds tolerable overhead to the ordinary deep learning system.
In cloud computing, a third party hosts a client's data, which raises privacy and security concerns. To maintain privacy, data should be encrypted by cryptographic techniques. However, encrypting the ...data makes it unsuitable for indexing and fast processing, as data needs to be decrypted to plain text before it can be further processed. Homomorphic encryption helps to overcome this shortcoming by allowing users to perform operations on encrypted data without decryption. Many academics have attempted to address the issue of data security, but none have addressed the issue of data privacy in cloud computing as thoroughly as this study has. This paper discusses the challenges involved in maintaining the privacy of cloud-based data and the techniques used to address these challenges. It was identified that homomorphic encryption is the best solution of all. This work also identified and compared the various homomorphic encryption schemes which are capable of ensuring the privacy of data in cloud storage and ways to implement them through libraries.
PDF
