Abstract:Highlights • We propose a lightweight computational scheme for private data aggregation. By adopting the CRT and encoding theory, private data is encoded into polynomial coefficients for computation and the random data are introduced to ensure that private data is not leaked to non-collusive servers. • We propose a non-interactive verifiable security aggregation protocol that satisfies the requirements of clients dropping out and partial server corruption. Furthermore, by introducing redundant shares and utilizing results returned by delayed servers, the proposed scheme achieves verifiable computation. • We provide a thorough analysis of the robustness, verification and security of our proposed. The experimental results show that the proposed scheme is effective in both computation and verification. Data aggregation involves the integration of relevant data generated across platforms and devices, leveraging the potential value of sensory data. However, in addition to security and efficiency, which are the basic requirements for data aggregation involving private data, how to achieve fault tolerance and interference of aggregation in real computing networks is imminent and is the main contribution of this paper. In this paper, we propose a secure aggregation framework involving multiple servers based on coding theory, which is not only robust to clients dropping out and tolerant to partial server withdrawal but also resistant to malicious computation by servers and forgery attacks by adversaries. In particular, the proposed protocol employs the Chinese Residual Theorem (CRT) to encode private data and constructs Lagrange interpolation polynomials to perform aggregation, which achieves lightweight privacy preservation while achieving robust, verifiable and secure aggregation goals.

Privacy-preserving Data Aggregation Without Secure Channel: Multivariate Polynomial Evaluation

Data Aggregation without Secure Channel: How to Evaluate a Multivariate Polynomial Securely

Collusion-Tolerable Privacy-Preserving Sum and Product Calculation without Secure Channel

Projected Federated Averaging with Heterogeneous Differential Privacy.

PPPA: A practical privacy-preserving aggregation scheme for smart grid communications

Privacy Preserving PCA for Multiparty Modeling

PANDA: Lightweight Non-Interactive Privacy-Preserving Data Aggregation for Constrained Devices

Privacy-Enhanced And Multifunctional Health Data Aggregation Under Differential Privacy Guarantees

Practical and Efficient Secure Aggregation for Privacy-Preserving Machine Learning.

PPCL: Privacy-preserving collaborative learning for mitigating indirect information leakage

Aggregate in My Way: Privacy-preserving Data Aggregation Without Trusted Authority in ICN

PDA: Semantically Secure Time-Series Data Analytics with Dynamic User Groups

Hiding in the Crowd: A Massively Distributed Algorithm for Private Averaging with Malicious Adversaries

Efficient and Secure Quantile Aggregation of Private Data Streams

Secure Single-Server Aggregation with (Poly)Logarithmic Overhead

RASE: Efficient Privacy-preserving Data Aggregation against Disclosure Attacks for IoTs

A privacy preserving quantum aggregating technique

Samplable Anonymous Aggregation for Private Federated Data Analysis

A coding computation scheme for secure aggregation

Ensuring Semantic Validity in Privacy-Preserving Aggregate Statistics

Bit-efficient Numerical Aggregation and Stronger Privacy for Trust in Federated Analytics